feat: support third part models of baidu (#1046)

百度千帆平台上的第三方大模型调用

.github/workflows/linux-release.yml

@@ -23,7 +23,7 @@ jobs:
       - uses: actions/setup-node@v3
         with:
           node-version: 16
-      - name: Build Frontend (theme default)
+      - name: Build Frontend
         env:
           CI: ""
         run: |
@@ -38,7 +38,7 @@ jobs:
       - name: Build Backend (amd64)
         run: |
           go mod download
-          go build -ldflags "-s -w -X 'one-api/common.Version=$(git describe --tags)' -extldflags '-static'" -o one-api
+          go build -ldflags "-s -w -X 'github.com/songquanpeng/one-api/common.Version=$(git describe --tags)' -extldflags '-static'" -o one-api
 
       - name: Build Backend (arm64)
         run: |

.github/workflows/macos-release.yml

@@ -23,7 +23,7 @@ jobs:
       - uses: actions/setup-node@v3
         with:
           node-version: 16
-      - name: Build Frontend (theme default)
+      - name: Build Frontend
         env:
           CI: ""
         run: |
@@ -38,7 +38,7 @@ jobs:
       - name: Build Backend
         run: |
           go mod download
-          go build -ldflags "-X 'one-api/common.Version=$(git describe --tags)'" -o one-api-macos
+          go build -ldflags "-X 'github.com/songquanpeng/one-api/common.Version=$(git describe --tags)'" -o one-api-macos
       - name: Release
         uses: softprops/action-gh-release@v1
         if: startsWith(github.ref, 'refs/tags/')

.github/workflows/windows-release.yml

@@ -26,7 +26,7 @@ jobs:
       - uses: actions/setup-node@v3
         with:
           node-version: 16
-      - name: Build Frontend (theme default)
+      - name: Build Frontend
         env:
           CI: ""
         run: |
@@ -41,7 +41,7 @@ jobs:
       - name: Build Backend
         run: |
           go mod download
-          go build -ldflags "-s -w -X 'one-api/common.Version=$(git describe --tags)'" -o one-api.exe
+          go build -ldflags "-s -w -X 'github.com/songquanpeng/one-api/common.Version=$(git describe --tags)'" -o one-api.exe
       - name: Release
         uses: softprops/action-gh-release@v1
         if: startsWith(github.ref, 'refs/tags/')

.gitignore

@@ -6,4 +6,5 @@ upload
 build
 *.db-journal
 logs
-data
+data
+/web/node_modules

Dockerfile

@@ -23,7 +23,7 @@ ADD go.mod go.sum ./
 RUN go mod download
 COPY . .
 COPY --from=builder /web/build ./web/build
-RUN go build -ldflags "-s -w -X 'one-api/common.Version=$(cat VERSION)' -extldflags '-static'" -o one-api
+RUN go build -ldflags "-s -w -X 'github.com/songquanpeng/one-api/common.Version=$(cat VERSION)' -extldflags '-static'" -o one-api
 
 FROM alpine
 

README.en.md

@@ -134,12 +134,12 @@ The initial account username is `root` and password is `123456`.
    git clone https://github.com/songquanpeng/one-api.git
    
    # Build the frontend
-   cd one-api/web
+   cd one-api/web/default
    npm install
    npm run build
    
    # Build the backend
-   cd ..
+   cd ../..
    go mod download
    go build -ldflags "-s -w" -o one-api
    ```

README.ja.md

@@ -135,12 +135,12 @@ sudo service nginx restart
    git clone https://github.com/songquanpeng/one-api.git
 
    # フロントエンドのビルド
-   cd one-api/web
+   cd one-api/web/default
    npm install
    npm run build
 
    # バックエンドのビルド
-   cd ..
+   cd ../..
    go mod download
    go build -ldflags "-s -w" -o one-api
    ```

README.md

@@ -67,12 +67,17 @@ _✨ 通过标准的 OpenAI API 格式访问所有的大模型，开箱即用
    + [x] [OpenAI ChatGPT 系列模型](https://platform.openai.com/docs/guides/gpt/chat-completions-api)（支持 [Azure OpenAI API](https://learn.microsoft.com/en-us/azure/ai-services/openai/reference)）
    + [x] [Anthropic Claude 系列模型](https://anthropic.com)
    + [x] [Google PaLM2/Gemini 系列模型](https://developers.generativeai.google)
+   + [x] [Mistral 系列模型](https://mistral.ai/)
    + [x] [百度文心一言系列模型](https://cloud.baidu.com/doc/WENXINWORKSHOP/index.html)
    + [x] [阿里通义千问系列模型](https://help.aliyun.com/document_detail/2400395.html)
    + [x] [讯飞星火认知大模型](https://www.xfyun.cn/doc/spark/Web.html)
    + [x] [智谱 ChatGLM 系列模型](https://bigmodel.cn)
    + [x] [360 智脑](https://ai.360.cn)
    + [x] [腾讯混元大模型](https://cloud.tencent.com/document/product/1729)
+   + [x] [Moonshot AI](https://platform.moonshot.cn/)
+   + [x] [百川大模型](https://platform.baichuan-ai.com)
+   + [ ] [字节云雀大模型](https://www.volcengine.com/product/ark) (WIP)
+   + [x] [MINIMAX](https://api.minimax.chat/)
 2. 支持配置镜像以及众多[第三方代理服务](https://iamazing.cn/page/openai-api-third-party-services)。
 3. 支持通过**负载均衡**的方式访问多个渠道。
 4. 支持 **stream 模式**，可以通过流式传输实现打字机效果。
@@ -174,12 +179,12 @@ docker-compose ps
    git clone https://github.com/songquanpeng/one-api.git
    
    # 构建前端
-   cd one-api/web
+   cd one-api/web/default
    npm install
    npm run build
    
    # 构建后端
-   cd ..
+   cd ../..
    go mod download
    go build -ldflags "-s -w" -o one-api
    ````

common/config/config.go

@@ -1,7 +1,7 @@
 package config
 
 import (
-	"one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/helper"
 	"os"
 	"strconv"
 	"sync"

common/constants.go

@@ -63,6 +63,10 @@ const (
 	ChannelTypeFastGPT        = 22
 	ChannelTypeTencent        = 23
 	ChannelTypeGemini         = 24
+	ChannelTypeMoonshot       = 25
+	ChannelTypeBaichuan       = 26
+	ChannelTypeMinimax        = 27
+	ChannelTypeMistral        = 28
 )
 
 var ChannelBaseURLs = []string{
@@ -91,4 +95,16 @@ var ChannelBaseURLs = []string{
 	"https://fastgpt.run/api/openapi",           // 22
 	"https://hunyuan.cloud.tencent.com",         // 23
 	"https://generativelanguage.googleapis.com", // 24
+	"https://api.moonshot.cn",                   // 25
+	"https://api.baichuan-ai.com",               // 26
+	"https://api.minimax.chat",                  // 27
+	"https://api.mistral.ai",                    // 28
 }
+
+const (
+	ConfigKeyPrefix = "cfg_"
+
+	ConfigKeyAPIVersion = ConfigKeyPrefix + "api_version"
+	ConfigKeyLibraryID  = ConfigKeyPrefix + "library_id"
+	ConfigKeyPlugin     = ConfigKeyPrefix + "plugin"
+)

common/database.go

@@ -1,6 +1,6 @@
 package common
 
-import "one-api/common/helper"
+import "github.com/songquanpeng/one-api/common/helper"
 
 var UsingSQLite = false
 var UsingPostgreSQL = false

common/email.go

@@ -5,8 +5,8 @@ import (
 	"crypto/tls"
 	"encoding/base64"
 	"fmt"
+	"github.com/songquanpeng/one-api/common/config"
 	"net/smtp"
-	"one-api/common/config"
 	"strings"
 	"time"
 )

common/embed-file-system.go

@@ -15,10 +15,7 @@ type embedFileSystem struct {
 
 func (e embedFileSystem) Exists(prefix string, path string) bool {
 	_, err := e.Open(path)
-	if err != nil {
-		return false
-	}
-	return true
+	return err == nil
 }
 
 func EmbedFolder(fsEmbed embed.FS, targetPath string) static.ServeFileSystem {

common/gin.go

@@ -8,12 +8,24 @@ import (
 	"strings"
 )
 
-func UnmarshalBodyReusable(c *gin.Context, v any) error {
+const KeyRequestBody = "key_request_body"
+
+func GetRequestBody(c *gin.Context) ([]byte, error) {
+	requestBody, _ := c.Get(KeyRequestBody)
+	if requestBody != nil {
+		return requestBody.([]byte), nil
+	}
 	requestBody, err := io.ReadAll(c.Request.Body)
 	if err != nil {
-		return err
+		return nil, err
 	}
-	err = c.Request.Body.Close()
+	_ = c.Request.Body.Close()
+	c.Set(KeyRequestBody, requestBody)
+	return requestBody.([]byte), nil
+}
+
+func UnmarshalBodyReusable(c *gin.Context, v any) error {
+	requestBody, err := GetRequestBody(c)
 	if err != nil {
 		return err
 	}

common/group-ratio.go

@@ -2,7 +2,7 @@ package common
 
 import (
 	"encoding/json"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/logger"
 )
 
 var GroupRatio = map[string]float64{

common/helper/helper.go

@@ -3,11 +3,11 @@ package helper
 import (
 	"fmt"
 	"github.com/google/uuid"
+	"github.com/songquanpeng/one-api/common/logger"
 	"html/template"
 	"log"
 	"math/rand"
 	"net"
-	"one-api/common/logger"
 	"os"
 	"os/exec"
 	"runtime"
@@ -107,13 +107,13 @@ func Seconds2Time(num int) (time string) {
 }
 
 func Interface2String(inter interface{}) string {
-	switch inter.(type) {
+	switch inter := inter.(type) {
 	case string:
-		return inter.(string)
+		return inter
 	case int:
-		return fmt.Sprintf("%d", inter.(int))
+		return fmt.Sprintf("%d", inter)
 	case float64:
-		return fmt.Sprintf("%f", inter.(float64))
+		return fmt.Sprintf("%f", inter)
 	}
 	return "Not Implemented"
 }
@@ -137,6 +137,7 @@ func GetUUID() string {
 }
 
 const keyChars = "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"
+const keyNumbers = "0123456789"
 
 func init() {
 	rand.Seed(time.Now().UnixNano())
@@ -168,6 +169,15 @@ func GetRandomString(length int) string {
 	return string(key)
 }
 
+func GetRandomNumberString(length int) string {
+	rand.Seed(time.Now().UnixNano())
+	key := make([]byte, length)
+	for i := 0; i < length; i++ {
+		key[i] = keyNumbers[rand.Intn(len(keyNumbers))]
+	}
+	return string(key)
+}
+
 func GetTimestamp() int64 {
 	return time.Now().Unix()
 }

common/image/image_test.go

@@ -12,7 +12,7 @@ import (
 	"strings"
 	"testing"
 
-	img "one-api/common/image"
+	img "github.com/songquanpeng/one-api/common/image"
 
 	"github.com/stretchr/testify/assert"
 	_ "golang.org/x/image/webp"

common/init.go

@@ -3,9 +3,9 @@ package common
 import (
 	"flag"
 	"fmt"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
 	"log"
-	"one-api/common/config"
-	"one-api/common/logger"
 	"os"
 	"path/filepath"
 )

common/logger/logger.go

@@ -13,6 +13,7 @@ import (
 )
 
 const (
+	loggerDEBUG = "DEBUG"
 	loggerINFO  = "INFO"
 	loggerWarn  = "WARN"
 	loggerError = "ERR"
@@ -55,6 +56,10 @@ func SysError(s string) {
 	_, _ = fmt.Fprintf(gin.DefaultErrorWriter, "[SYS] %v | %s \n", t.Format("2006/01/02 - 15:04:05"), s)
 }
 
+func Debug(ctx context.Context, msg string) {
+	logHelper(ctx, loggerDEBUG, msg)
+}
+
 func Info(ctx context.Context, msg string) {
 	logHelper(ctx, loggerINFO, msg)
 }
@@ -67,16 +72,20 @@ func Error(ctx context.Context, msg string) {
 	logHelper(ctx, loggerError, msg)
 }
 
+func Debugf(ctx context.Context, format string, a ...any) {
+	Debug(ctx, fmt.Sprintf(format, a...))
+}
+
 func Infof(ctx context.Context, format string, a ...any) {
-	Info(ctx, fmt.Sprintf(format, a))
+	Info(ctx, fmt.Sprintf(format, a...))
 }
 
 func Warnf(ctx context.Context, format string, a ...any) {
-	Warn(ctx, fmt.Sprintf(format, a))
+	Warn(ctx, fmt.Sprintf(format, a...))
 }
 
 func Errorf(ctx context.Context, format string, a ...any) {
-	Error(ctx, fmt.Sprintf(format, a))
+	Error(ctx, fmt.Sprintf(format, a...))
 }
 
 func logHelper(ctx context.Context, level string, msg string) {

common/model-ratio.go

@@ -2,92 +2,86 @@ package common
 
 import (
 	"encoding/json"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/logger"
 	"strings"
 	"time"
 )
 
-var DalleSizeRatios = map[string]map[string]float64{
-	"dall-e-2": {
-		"256x256":   1,
-		"512x512":   1.125,
-		"1024x1024": 1.25,
-	},
-	"dall-e-3": {
-		"1024x1024": 1,
-		"1024x1792": 2,
-		"1792x1024": 2,
-	},
-}
-
-var DalleGenerationImageAmounts = map[string][2]int{
-	"dall-e-2": {1, 10},
-	"dall-e-3": {1, 1}, // OpenAI allows n=1 currently.
-}
-
-var DalleImagePromptLengthLimitations = map[string]int{
-	"dall-e-2": 1000,
-	"dall-e-3": 4000,
-}
+const (
+	USD2RMB = 7
+	USD     = 500 // $0.002 = 1 -> $1 = 500
+	RMB     = USD / USD2RMB
+)
 
 // ModelRatio
 // https://platform.openai.com/docs/models/model-endpoint-compatibility
 // https://cloud.baidu.com/doc/WENXINWORKSHOP/s/Blfmc9dlf
 // https://openai.com/pricing
-// TODO: when a new api is enabled, check the pricing here
 // 1 === $0.002 / 1K tokens
 // 1 === ￥0.014 / 1k tokens
 var ModelRatio = map[string]float64{
-	"gpt-4":                     15,
-	"gpt-4-0314":                15,
-	"gpt-4-0613":                15,
-	"gpt-4-32k":                 30,
-	"gpt-4-32k-0314":            30,
-	"gpt-4-32k-0613":            30,
-	"gpt-4-1106-preview":        5,    // $0.01 / 1K tokens
-	"gpt-4-vision-preview":      5,    // $0.01 / 1K tokens
-	"gpt-3.5-turbo":             0.75, // $0.0015 / 1K tokens
-	"gpt-3.5-turbo-0301":        0.75,
-	"gpt-3.5-turbo-0613":        0.75,
-	"gpt-3.5-turbo-16k":         1.5, // $0.003 / 1K tokens
-	"gpt-3.5-turbo-16k-0613":    1.5,
-	"gpt-3.5-turbo-instruct":    0.75, // $0.0015 / 1K tokens
-	"gpt-3.5-turbo-1106":        0.5,  // $0.001 / 1K tokens
-	"davinci-002":               1,    // $0.002 / 1K tokens
-	"babbage-002":               0.2,  // $0.0004 / 1K tokens
-	"text-ada-001":              0.2,
-	"text-babbage-001":          0.25,
-	"text-curie-001":            1,
-	"text-davinci-002":          10,
-	"text-davinci-003":          10,
-	"text-davinci-edit-001":     10,
-	"code-davinci-edit-001":     10,
-	"whisper-1":                 15,  // $0.006 / minute -> $0.006 / 150 words -> $0.006 / 200 tokens -> $0.03 / 1k tokens
-	"tts-1":                     7.5, // $0.015 / 1K characters
-	"tts-1-1106":                7.5,
-	"tts-1-hd":                  15, // $0.030 / 1K characters
-	"tts-1-hd-1106":             15,
-	"davinci":                   10,
-	"curie":                     10,
-	"babbage":                   10,
-	"ada":                       10,
-	"text-embedding-ada-002":    0.05,
-	"text-search-ada-doc-001":   10,
-	"text-moderation-stable":    0.1,
-	"text-moderation-latest":    0.1,
-	"dall-e-2":                  8,      // $0.016 - $0.020 / image
-	"dall-e-3":                  20,     // $0.040 - $0.120 / image
-	"claude-instant-1":          0.815,  // $1.63 / 1M tokens
-	"claude-2":                  5.51,   // $11.02 / 1M tokens
-	"claude-2.0":                5.51,   // $11.02 / 1M tokens
-	"claude-2.1":                5.51,   // $11.02 / 1M tokens
-	"ERNIE-Bot":                 0.8572, // ￥0.012 / 1k tokens
-	"ERNIE-Bot-turbo":           0.5715, // ￥0.008 / 1k tokens
-	"ERNIE-Bot-4":               8.572,  // ￥0.12 / 1k tokens
-	"Embedding-V1":              0.1429, // ￥0.002 / 1k tokens
-	"PaLM-2":                    1,
-	"gemini-pro":                1,      // $0.00025 / 1k characters -> $0.001 / 1k tokens
-	"gemini-pro-vision":         1,      // $0.00025 / 1k characters -> $0.001 / 1k tokens
+	// https://openai.com/pricing
+	"gpt-4":                   15,
+	"gpt-4-0314":              15,
+	"gpt-4-0613":              15,
+	"gpt-4-32k":               30,
+	"gpt-4-32k-0314":          30,
+	"gpt-4-32k-0613":          30,
+	"gpt-4-1106-preview":      5,    // $0.01 / 1K tokens
+	"gpt-4-0125-preview":      5,    // $0.01 / 1K tokens
+	"gpt-4-turbo-preview":     5,    // $0.01 / 1K tokens
+	"gpt-4-vision-preview":    5,    // $0.01 / 1K tokens
+	"gpt-3.5-turbo":           0.75, // $0.0015 / 1K tokens
+	"gpt-3.5-turbo-0301":      0.75,
+	"gpt-3.5-turbo-0613":      0.75,
+	"gpt-3.5-turbo-16k":       1.5, // $0.003 / 1K tokens
+	"gpt-3.5-turbo-16k-0613":  1.5,
+	"gpt-3.5-turbo-instruct":  0.75, // $0.0015 / 1K tokens
+	"gpt-3.5-turbo-1106":      0.5,  // $0.001 / 1K tokens
+	"gpt-3.5-turbo-0125":      0.25, // $0.0005 / 1K tokens
+	"davinci-002":             1,    // $0.002 / 1K tokens
+	"babbage-002":             0.2,  // $0.0004 / 1K tokens
+	"text-ada-001":            0.2,
+	"text-babbage-001":        0.25,
+	"text-curie-001":          1,
+	"text-davinci-002":        10,
+	"text-davinci-003":        10,
+	"text-davinci-edit-001":   10,
+	"code-davinci-edit-001":   10,
+	"whisper-1":               15,  // $0.006 / minute -> $0.006 / 150 words -> $0.006 / 200 tokens -> $0.03 / 1k tokens
+	"tts-1":                   7.5, // $0.015 / 1K characters
+	"tts-1-1106":              7.5,
+	"tts-1-hd":                15, // $0.030 / 1K characters
+	"tts-1-hd-1106":           15,
+	"davinci":                 10,
+	"curie":                   10,
+	"babbage":                 10,
+	"ada":                     10,
+	"text-embedding-ada-002":  0.05,
+	"text-embedding-3-small":  0.01,
+	"text-embedding-3-large":  0.065,
+	"text-search-ada-doc-001": 10,
+	"text-moderation-stable":  0.1,
+	"text-moderation-latest":  0.1,
+	"dall-e-2":                8,     // $0.016 - $0.020 / image
+	"dall-e-3":                20,    // $0.040 - $0.120 / image
+	"claude-instant-1":        0.815, // $1.63 / 1M tokens
+	"claude-2":                5.51,  // $11.02 / 1M tokens
+	"claude-2.0":              5.51,  // $11.02 / 1M tokens
+	"claude-2.1":              5.51,  // $11.02 / 1M tokens
+	// https://cloud.baidu.com/doc/WENXINWORKSHOP/s/hlrk4akp7
+	"ERNIE-Bot":         0.8572,     // ￥0.012 / 1k tokens
+	"ERNIE-Bot-turbo":   0.5715,     // ￥0.008 / 1k tokens
+	"ERNIE-Bot-4":       0.12 * RMB, // ￥0.12 / 1k tokens
+	"ERNIE-Bot-8k":      0.024 * RMB,
+	"Embedding-V1":      0.1429, // ￥0.002 / 1k tokens
+	"PaLM-2":            1,
+	"gemini-pro":        1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
+	"gemini-pro-vision": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
+	// https://open.bigmodel.cn/pricing
+	"glm-4":                     0.1 * RMB,
+	"glm-4v":                    0.1 * RMB,
+	"glm-3-turbo":               0.005 * RMB,
 	"chatglm_turbo":             0.3572, // ￥0.005 / 1k tokens
 	"chatglm_pro":               0.7143, // ￥0.01 / 1k tokens
 	"chatglm_std":               0.3572, // ￥0.005 / 1k tokens
@@ -98,11 +92,52 @@ var ModelRatio = map[string]float64{
 	"qwen-max-longcontext":      1.4286, // ￥0.02 / 1k tokens
 	"text-embedding-v1":         0.05,   // ￥0.0007 / 1k tokens
 	"SparkDesk":                 1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v1.1":            1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v2.1":            1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v3.1":            1.2858, // ￥0.018 / 1k tokens
+	"SparkDesk-v3.5":            1.2858, // ￥0.018 / 1k tokens
 	"360GPT_S2_V9":              0.8572, // ¥0.012 / 1k tokens
 	"embedding-bert-512-v1":     0.0715, // ¥0.001 / 1k tokens
 	"embedding_s1_v1":           0.0715, // ¥0.001 / 1k tokens
 	"semantic_similarity_s1_v1": 0.0715, // ¥0.001 / 1k tokens
 	"hunyuan":                   7.143,  // ¥0.1 / 1k tokens  // https://cloud.tencent.com/document/product/1729/97731#e0e6be58-60c8-469f-bdeb-6c264ce3b4d0
+	"ChatStd":                   0.01 * RMB,
+	"ChatPro":                   0.1 * RMB,
+	// https://platform.moonshot.cn/pricing
+	"moonshot-v1-8k":   0.012 * RMB,
+	"moonshot-v1-32k":  0.024 * RMB,
+	"moonshot-v1-128k": 0.06 * RMB,
+	// https://platform.baichuan-ai.com/price
+	"Baichuan2-Turbo":      0.008 * RMB,
+	"Baichuan2-Turbo-192k": 0.016 * RMB,
+	"Baichuan2-53B":        0.02 * RMB,
+	// https://api.minimax.chat/document/price
+	"abab6-chat":    0.1 * RMB,
+	"abab5.5-chat":  0.015 * RMB,
+	"abab5.5s-chat": 0.005 * RMB,
+	// https://docs.mistral.ai/platform/pricing/
+	"open-mistral-7b":       0.25 / 1000 * USD,
+	"open-mixtral-8x7b":     0.7 / 1000 * USD,
+	"mistral-small-latest":  2.0 / 1000 * USD,
+	"mistral-medium-latest": 2.7 / 1000 * USD,
+	"mistral-large-latest":  8.0 / 1000 * USD,
+	"mistral-embed":         0.1 / 1000 * USD,
+}
+
+var CompletionRatio = map[string]float64{}
+
+var DefaultModelRatio map[string]float64
+var DefaultCompletionRatio map[string]float64
+
+func init() {
+	DefaultModelRatio = make(map[string]float64)
+	for k, v := range ModelRatio {
+		DefaultModelRatio[k] = v
+	}
+	DefaultCompletionRatio = make(map[string]float64)
+	for k, v := range CompletionRatio {
+		DefaultCompletionRatio[k] = v
+	}
 }
 
 func ModelRatio2JSONString() string {
@@ -123,6 +158,9 @@ func GetModelRatio(name string) float64 {
 		name = strings.TrimSuffix(name, "-internet")
 	}
 	ratio, ok := ModelRatio[name]
+	if !ok {
+		ratio, ok = DefaultModelRatio[name]
+	}
 	if !ok {
 		logger.SysError("model ratio not found: " + name)
 		return 30
@@ -130,8 +168,32 @@ func GetModelRatio(name string) float64 {
 	return ratio
 }
 
+func CompletionRatio2JSONString() string {
+	jsonBytes, err := json.Marshal(CompletionRatio)
+	if err != nil {
+		logger.SysError("error marshalling completion ratio: " + err.Error())
+	}
+	return string(jsonBytes)
+}
+
+func UpdateCompletionRatioByJSONString(jsonStr string) error {
+	CompletionRatio = make(map[string]float64)
+	return json.Unmarshal([]byte(jsonStr), &CompletionRatio)
+}
+
 func GetCompletionRatio(name string) float64 {
+	if ratio, ok := CompletionRatio[name]; ok {
+		return ratio
+	}
+	if ratio, ok := DefaultCompletionRatio[name]; ok {
+		return ratio
+	}
 	if strings.HasPrefix(name, "gpt-3.5") {
+		if strings.HasSuffix(name, "0125") {
+			// https://openai.com/blog/new-embedding-models-and-api-updates
+			// Updated GPT-3.5 Turbo model and lower pricing
+			return 3
+		}
 		if strings.HasSuffix(name, "1106") {
 			return 2
 		}
@@ -158,5 +220,8 @@ func GetCompletionRatio(name string) float64 {
 	if strings.HasPrefix(name, "claude-2") {
 		return 2.965517
 	}
+	if strings.HasPrefix(name, "mistral-") {
+		return 3
+	}
 	return 1
 }

common/random.go

@@ -0,0 +1,8 @@
+package common
+
+import "math/rand"
+
+// RandRange returns a random number between min and max (max is not included)
+func RandRange(min, max int) int {
+	return min + rand.Intn(max-min)
+}

common/redis.go

@@ -3,7 +3,7 @@ package common
 import (
 	"context"
 	"github.com/go-redis/redis/v8"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/logger"
 	"os"
 	"time"
 )

common/utils.go

@@ -2,7 +2,7 @@ package common
 
 import (
 	"fmt"
-	"one-api/common/config"
+	"github.com/songquanpeng/one-api/common/config"
 )
 
 func LogQuota(quota int) string {

controller/billing.go

@@ -2,9 +2,9 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
-	"one-api/common/config"
-	"one-api/model"
-	"one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/model"
+	relaymodel "github.com/songquanpeng/one-api/relay/model"
 )
 
 func GetSubscription(c *gin.Context) {
@@ -22,13 +22,15 @@ func GetSubscription(c *gin.Context) {
 	} else {
 		userId := c.GetInt("id")
 		remainQuota, err = model.GetUserQuota(userId)
-		usedQuota, err = model.GetUserUsedQuota(userId)
+		if err != nil {
+			usedQuota, err = model.GetUserUsedQuota(userId)
+		}
 	}
 	if expiredTime <= 0 {
 		expiredTime = 0
 	}
 	if err != nil {
-		Error := openai.Error{
+		Error := relaymodel.Error{
 			Message: err.Error(),
 			Type:    "upstream_error",
 		}
@@ -70,7 +72,7 @@ func GetUsage(c *gin.Context) {
 		quota, err = model.GetUserUsedQuota(userId)
 	}
 	if err != nil {
-		Error := openai.Error{
+		Error := relaymodel.Error{
 			Message: err.Error(),
 			Type:    "one_api_error",
 		}

controller/channel-billing.go

@@ -4,13 +4,13 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/relay/util"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/logger"
-	"one-api/model"
-	"one-api/relay/util"
 	"strconv"
 	"time"
 

controller/channel-test.go

@@ -5,102 +5,34 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/middleware"
+	"github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/helper"
+	relaymodel "github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/logger"
-	"one-api/model"
-	"one-api/relay/channel/openai"
-	"one-api/relay/util"
+	"net/http/httptest"
+	"net/url"
 	"strconv"
+	"strings"
 	"sync"
 	"time"
 
 	"github.com/gin-gonic/gin"
 )
 
-func testChannel(channel *model.Channel, request openai.ChatRequest) (err error, openaiErr *openai.Error) {
-	switch channel.Type {
-	case common.ChannelTypePaLM:
-		fallthrough
-	case common.ChannelTypeGemini:
-		fallthrough
-	case common.ChannelTypeAnthropic:
-		fallthrough
-	case common.ChannelTypeBaidu:
-		fallthrough
-	case common.ChannelTypeZhipu:
-		fallthrough
-	case common.ChannelTypeAli:
-		fallthrough
-	case common.ChannelType360:
-		fallthrough
-	case common.ChannelTypeXunfei:
-		return errors.New("该渠道类型当前版本不支持测试，请手动测试"), nil
-	case common.ChannelTypeAzure:
-		request.Model = "gpt-35-turbo"
-		defer func() {
-			if err != nil {
-				err = errors.New("请确保已在 Azure 上创建了 gpt-35-turbo 模型，并且 apiVersion 已正确填写！")
-			}
-		}()
-	default:
-		request.Model = "gpt-3.5-turbo"
-	}
-	requestURL := common.ChannelBaseURLs[channel.Type]
-	if channel.Type == common.ChannelTypeAzure {
-		requestURL = util.GetFullRequestURL(channel.GetBaseURL(), fmt.Sprintf("/openai/deployments/%s/chat/completions?api-version=2023-03-15-preview", request.Model), channel.Type)
-	} else {
-		if baseURL := channel.GetBaseURL(); len(baseURL) > 0 {
-			requestURL = baseURL
-		}
-
-		requestURL = util.GetFullRequestURL(requestURL, "/v1/chat/completions", channel.Type)
-	}
-	jsonData, err := json.Marshal(request)
-	if err != nil {
-		return err, nil
-	}
-	req, err := http.NewRequest("POST", requestURL, bytes.NewBuffer(jsonData))
-	if err != nil {
-		return err, nil
-	}
-	if channel.Type == common.ChannelTypeAzure {
-		req.Header.Set("api-key", channel.Key)
-	} else {
-		req.Header.Set("Authorization", "Bearer "+channel.Key)
-	}
-	req.Header.Set("Content-Type", "application/json")
-	resp, err := util.HTTPClient.Do(req)
-	if err != nil {
-		return err, nil
-	}
-	defer resp.Body.Close()
-	var response openai.SlimTextResponse
-	body, err := io.ReadAll(resp.Body)
-	if err != nil {
-		return err, nil
-	}
-	err = json.Unmarshal(body, &response)
-	if err != nil {
-		return fmt.Errorf("Error: %s\nResp body: %s", err, body), nil
-	}
-	if response.Usage.CompletionTokens == 0 {
-		if response.Error.Message == "" {
-			response.Error.Message = "补全 tokens 非预期返回 0"
-		}
-		return errors.New(fmt.Sprintf("type %s, code %v, message %s", response.Error.Type, response.Error.Code, response.Error.Message)), &response.Error
-	}
-	return nil, nil
-}
-
-func buildTestRequest() *openai.ChatRequest {
-	testRequest := &openai.ChatRequest{
-		Model:     "", // this will be set later
+func buildTestRequest() *relaymodel.GeneralOpenAIRequest {
+	testRequest := &relaymodel.GeneralOpenAIRequest{
 		MaxTokens: 1,
+		Stream:    false,
+		Model:     "gpt-3.5-turbo",
 	}
-	testMessage := openai.Message{
+	testMessage := relaymodel.Message{
 		Role:    "user",
 		Content: "hi",
 	}
@@ -108,6 +40,72 @@ func buildTestRequest() *openai.ChatRequest {
 	return testRequest
 }
 
+func testChannel(channel *model.Channel) (err error, openaiErr *relaymodel.Error) {
+	w := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(w)
+	c.Request = &http.Request{
+		Method: "POST",
+		URL:    &url.URL{Path: "/v1/chat/completions"},
+		Body:   nil,
+		Header: make(http.Header),
+	}
+	c.Request.Header.Set("Authorization", "Bearer "+channel.Key)
+	c.Request.Header.Set("Content-Type", "application/json")
+	c.Set("channel", channel.Type)
+	c.Set("base_url", channel.GetBaseURL())
+	middleware.SetupContextForSelectedChannel(c, channel, "")
+	meta := util.GetRelayMeta(c)
+	apiType := constant.ChannelType2APIType(channel.Type)
+	adaptor := helper.GetAdaptor(apiType)
+	if adaptor == nil {
+		return fmt.Errorf("invalid api type: %d, adaptor is nil", apiType), nil
+	}
+	adaptor.Init(meta)
+	modelName := adaptor.GetModelList()[0]
+	if !strings.Contains(channel.Models, modelName) {
+		modelNames := strings.Split(channel.Models, ",")
+		if len(modelNames) > 0 {
+			modelName = modelNames[0]
+		}
+	}
+	request := buildTestRequest()
+	request.Model = modelName
+	meta.OriginModelName, meta.ActualModelName = modelName, modelName
+	convertedRequest, err := adaptor.ConvertRequest(c, constant.RelayModeChatCompletions, request)
+	if err != nil {
+		return err, nil
+	}
+	jsonData, err := json.Marshal(convertedRequest)
+	if err != nil {
+		return err, nil
+	}
+	requestBody := bytes.NewBuffer(jsonData)
+	c.Request.Body = io.NopCloser(requestBody)
+	resp, err := adaptor.DoRequest(c, meta, requestBody)
+	if err != nil {
+		return err, nil
+	}
+	if resp.StatusCode != http.StatusOK {
+		err := util.RelayErrorHandler(resp)
+		return fmt.Errorf("status code %d: %s", resp.StatusCode, err.Error.Message), &err.Error
+	}
+	usage, respErr := adaptor.DoResponse(c, resp, meta)
+	if respErr != nil {
+		return fmt.Errorf("%s", respErr.Error.Message), &respErr.Error
+	}
+	if usage == nil {
+		return errors.New("usage is nil"), nil
+	}
+	result := w.Result()
+	// print result.Body
+	respBody, err := io.ReadAll(result.Body)
+	if err != nil {
+		return err, nil
+	}
+	logger.SysLog(fmt.Sprintf("testing channel #%d, response: \n%s", channel.Id, string(respBody)))
+	return nil, nil
+}
+
 func TestChannel(c *gin.Context) {
 	id, err := strconv.Atoi(c.Param("id"))
 	if err != nil {
@@ -125,9 +123,8 @@ func TestChannel(c *gin.Context) {
 		})
 		return
 	}
-	testRequest := buildTestRequest()
 	tik := time.Now()
-	err, _ = testChannel(channel, *testRequest)
+	err, _ = testChannel(channel)
 	tok := time.Now()
 	milliseconds := tok.Sub(tik).Milliseconds()
 	go channel.UpdateResponseTime(milliseconds)
@@ -192,7 +189,6 @@ func testAllChannels(notify bool) error {
 	if err != nil {
 		return err
 	}
-	testRequest := buildTestRequest()
 	var disableThreshold = int64(config.ChannelDisableThreshold * 1000)
 	if disableThreshold == 0 {
 		disableThreshold = 10000000 // a impossible value
@@ -201,7 +197,7 @@ func testAllChannels(notify bool) error {
 		for _, channel := range channels {
 			isChannelEnabled := channel.Status == common.ChannelStatusEnabled
 			tik := time.Now()
-			err, openaiErr := testChannel(channel, *testRequest)
+			err, openaiErr := testChannel(channel)
 			tok := time.Now()
 			milliseconds := tok.Sub(tik).Milliseconds()
 			if isChannelEnabled && milliseconds > disableThreshold {

controller/channel.go

@@ -2,10 +2,10 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/model"
 	"strconv"
 	"strings"
 )

controller/github.go

@@ -7,12 +7,12 @@ import (
 	"fmt"
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
-	"one-api/model"
 	"strconv"
 	"time"
 )

controller/group.go

@@ -2,13 +2,13 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
 	"net/http"
-	"one-api/common"
 )
 
 func GetGroups(c *gin.Context) {
 	groupNames := make([]string, 0)
-	for groupName, _ := range common.GroupRatio {
+	for groupName := range common.GroupRatio {
 		groupNames = append(groupNames, groupName)
 	}
 	c.JSON(http.StatusOK, gin.H{

controller/log.go

@@ -2,9 +2,9 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common/config"
-	"one-api/model"
 	"strconv"
 )
 

controller/misc.go

@@ -3,10 +3,10 @@ package controller
 import (
 	"encoding/json"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/model"
 	"strings"
 
 	"github.com/gin-gonic/gin"

controller/model.go

@@ -3,7 +3,14 @@ package controller
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/channel/ai360"
+	"github.com/songquanpeng/one-api/relay/channel/baichuan"
+	"github.com/songquanpeng/one-api/relay/channel/minimax"
+	"github.com/songquanpeng/one-api/relay/channel/mistral"
+	"github.com/songquanpeng/one-api/relay/channel/moonshot"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/helper"
+	relaymodel "github.com/songquanpeng/one-api/relay/model"
 )
 
 // https://platform.openai.com/docs/api-reference/models/list
@@ -53,547 +60,79 @@ func init() {
 		IsBlocking:         false,
 	})
 	// https://platform.openai.com/docs/models/model-endpoint-compatibility
-	openAIModels = []OpenAIModels{
-		{
-			Id:         "dall-e-2",
+	for i := 0; i < constant.APITypeDummy; i++ {
+		if i == constant.APITypeAIProxyLibrary {
+			continue
+		}
+		adaptor := helper.GetAdaptor(i)
+		channelName := adaptor.GetChannelName()
+		modelNames := adaptor.GetModelList()
+		for _, modelName := range modelNames {
+			openAIModels = append(openAIModels, OpenAIModels{
+				Id:         modelName,
+				Object:     "model",
+				Created:    1626777600,
+				OwnedBy:    channelName,
+				Permission: permission,
+				Root:       modelName,
+				Parent:     nil,
+			})
+		}
+	}
+	for _, modelName := range ai360.ModelList {
+		openAIModels = append(openAIModels, OpenAIModels{
+			Id:         modelName,
 			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "dall-e-2",
-			Parent:     nil,
-		},
-		{
-			Id:         "dall-e-3",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "dall-e-3",
-			Parent:     nil,
-		},
-		{
-			Id:         "whisper-1",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "whisper-1",
-			Parent:     nil,
-		},
-		{
-			Id:         "tts-1",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "tts-1",
-			Parent:     nil,
-		},
-		{
-			Id:         "tts-1-1106",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "tts-1-1106",
-			Parent:     nil,
-		},
-		{
-			Id:         "tts-1-hd",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "tts-1-hd",
-			Parent:     nil,
-		},
-		{
-			Id:         "tts-1-hd-1106",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "tts-1-hd-1106",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo-0301",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo-0301",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo-0613",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo-0613",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo-16k",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo-16k",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo-16k-0613",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo-16k-0613",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo-1106",
-			Object:     "model",
-			Created:    1699593571,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo-1106",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-3.5-turbo-instruct",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-3.5-turbo-instruct",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-0314",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-0314",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-0613",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-0613",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-32k",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-32k",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-32k-0314",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-32k-0314",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-32k-0613",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-32k-0613",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-1106-preview",
-			Object:     "model",
-			Created:    1699593571,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-1106-preview",
-			Parent:     nil,
-		},
-		{
-			Id:         "gpt-4-vision-preview",
-			Object:     "model",
-			Created:    1699593571,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "gpt-4-vision-preview",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-embedding-ada-002",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-embedding-ada-002",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-davinci-003",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-davinci-003",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-davinci-002",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-davinci-002",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-curie-001",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-curie-001",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-babbage-001",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-babbage-001",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-ada-001",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-ada-001",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-moderation-latest",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-moderation-latest",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-moderation-stable",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-moderation-stable",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-davinci-edit-001",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "text-davinci-edit-001",
-			Parent:     nil,
-		},
-		{
-			Id:         "code-davinci-edit-001",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "code-davinci-edit-001",
-			Parent:     nil,
-		},
-		{
-			Id:         "davinci-002",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "davinci-002",
-			Parent:     nil,
-		},
-		{
-			Id:         "babbage-002",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "openai",
-			Permission: permission,
-			Root:       "babbage-002",
-			Parent:     nil,
-		},
-		{
-			Id:         "claude-instant-1",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "anthropic",
-			Permission: permission,
-			Root:       "claude-instant-1",
-			Parent:     nil,
-		},
-		{
-			Id:         "claude-2",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "anthropic",
-			Permission: permission,
-			Root:       "claude-2",
-			Parent:     nil,
-		},
-		{
-			Id:         "claude-2.1",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "anthropic",
-			Permission: permission,
-			Root:       "claude-2.1",
-			Parent:     nil,
-		},
-		{
-			Id:         "claude-2.0",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "anthropic",
-			Permission: permission,
-			Root:       "claude-2.0",
-			Parent:     nil,
-		},
-		{
-			Id:         "ERNIE-Bot",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "baidu",
-			Permission: permission,
-			Root:       "ERNIE-Bot",
-			Parent:     nil,
-		},
-		{
-			Id:         "ERNIE-Bot-turbo",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "baidu",
-			Permission: permission,
-			Root:       "ERNIE-Bot-turbo",
-			Parent:     nil,
-		},
-		{
-			Id:         "ERNIE-Bot-4",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "baidu",
-			Permission: permission,
-			Root:       "ERNIE-Bot-4",
-			Parent:     nil,
-		},
-		{
-			Id:         "Embedding-V1",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "baidu",
-			Permission: permission,
-			Root:       "Embedding-V1",
-			Parent:     nil,
-		},
-		{
-			Id:         "PaLM-2",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "google palm",
-			Permission: permission,
-			Root:       "PaLM-2",
-			Parent:     nil,
-		},
-		{
-			Id:         "gemini-pro",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "google gemini",
-			Permission: permission,
-			Root:       "gemini-pro",
-			Parent:     nil,
-		},
-		{
-			Id:         "gemini-pro-vision",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "google gemini",
-			Permission: permission,
-			Root:       "gemini-pro-vision",
-			Parent:     nil,
-		},
-		{
-			Id:         "chatglm_turbo",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "zhipu",
-			Permission: permission,
-			Root:       "chatglm_turbo",
-			Parent:     nil,
-		},
-		{
-			Id:         "chatglm_pro",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "zhipu",
-			Permission: permission,
-			Root:       "chatglm_pro",
-			Parent:     nil,
-		},
-		{
-			Id:         "chatglm_std",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "zhipu",
-			Permission: permission,
-			Root:       "chatglm_std",
-			Parent:     nil,
-		},
-		{
-			Id:         "chatglm_lite",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "zhipu",
-			Permission: permission,
-			Root:       "chatglm_lite",
-			Parent:     nil,
-		},
-		{
-			Id:         "qwen-turbo",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "ali",
-			Permission: permission,
-			Root:       "qwen-turbo",
-			Parent:     nil,
-		},
-		{
-			Id:         "qwen-plus",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "ali",
-			Permission: permission,
-			Root:       "qwen-plus",
-			Parent:     nil,
-		},
-		{
-			Id:         "qwen-max",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "ali",
-			Permission: permission,
-			Root:       "qwen-max",
-			Parent:     nil,
-		},
-		{
-			Id:         "qwen-max-longcontext",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "ali",
-			Permission: permission,
-			Root:       "qwen-max-longcontext",
-			Parent:     nil,
-		},
-		{
-			Id:         "text-embedding-v1",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "ali",
-			Permission: permission,
-			Root:       "text-embedding-v1",
-			Parent:     nil,
-		},
-		{
-			Id:         "SparkDesk",
-			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "xunfei",
-			Permission: permission,
-			Root:       "SparkDesk",
-			Parent:     nil,
-		},
-		{
-			Id:         "360GPT_S2_V9",
-			Object:     "model",
-			Created:    1677649963,
+			Created:    1626777600,
 			OwnedBy:    "360",
 			Permission: permission,
-			Root:       "360GPT_S2_V9",
+			Root:       modelName,
 			Parent:     nil,
-		},
-		{
-			Id:         "embedding-bert-512-v1",
+		})
+	}
+	for _, modelName := range moonshot.ModelList {
+		openAIModels = append(openAIModels, OpenAIModels{
+			Id:         modelName,
 			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "360",
+			Created:    1626777600,
+			OwnedBy:    "moonshot",
 			Permission: permission,
-			Root:       "embedding-bert-512-v1",
+			Root:       modelName,
 			Parent:     nil,
-		},
-		{
-			Id:         "embedding_s1_v1",
+		})
+	}
+	for _, modelName := range baichuan.ModelList {
+		openAIModels = append(openAIModels, OpenAIModels{
+			Id:         modelName,
 			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "360",
+			Created:    1626777600,
+			OwnedBy:    "baichuan",
 			Permission: permission,
-			Root:       "embedding_s1_v1",
+			Root:       modelName,
 			Parent:     nil,
-		},
-		{
-			Id:         "semantic_similarity_s1_v1",
+		})
+	}
+	for _, modelName := range minimax.ModelList {
+		openAIModels = append(openAIModels, OpenAIModels{
+			Id:         modelName,
 			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "360",
+			Created:    1626777600,
+			OwnedBy:    "minimax",
 			Permission: permission,
-			Root:       "semantic_similarity_s1_v1",
+			Root:       modelName,
 			Parent:     nil,
-		},
-		{
-			Id:         "hunyuan",
+		})
+	}
+	for _, modelName := range mistral.ModelList {
+		openAIModels = append(openAIModels, OpenAIModels{
+			Id:         modelName,
 			Object:     "model",
-			Created:    1677649963,
-			OwnedBy:    "tencent",
+			Created:    1626777600,
+			OwnedBy:    "mistralai",
 			Permission: permission,
-			Root:       "hunyuan",
+			Root:       modelName,
 			Parent:     nil,
-		},
+		})
 	}
 	openAIModelsMap = make(map[string]OpenAIModels)
 	for _, model := range openAIModels {
@@ -613,7 +152,7 @@ func RetrieveModel(c *gin.Context) {
 	if model, ok := openAIModelsMap[modelId]; ok {
 		c.JSON(200, model)
 	} else {
-		Error := openai.Error{
+		Error := relaymodel.Error{
 			Message: fmt.Sprintf("The model '%s' does not exist", modelId),
 			Type:    "invalid_request_error",
 			Param:   "model",

controller/option.go

@@ -2,10 +2,10 @@ package controller
 
 import (
 	"encoding/json"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/model"
 	"strings"
 
 	"github.com/gin-gonic/gin"

controller/redemption.go

@@ -2,10 +2,10 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/model"
 	"strconv"
 )
 

controller/relay.go

@@ -1,24 +1,28 @@
 package controller
 
 import (
+	"bytes"
+	"context"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/middleware"
+	dbmodel "github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/controller"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
-	"one-api/relay/constant"
-	"one-api/relay/controller"
-	"one-api/relay/util"
-	"strconv"
 )
 
 // https://platform.openai.com/docs/api-reference/chat
 
-func Relay(c *gin.Context) {
-	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
-	var err *openai.ErrorWithStatusCode
+func relay(c *gin.Context, relayMode int) *model.ErrorWithStatusCode {
+	var err *model.ErrorWithStatusCode
 	switch relayMode {
 	case constant.RelayModeImagesGenerations:
 		err = controller.RelayImageHelper(c, relayMode)
@@ -29,39 +33,96 @@ func Relay(c *gin.Context) {
 	case constant.RelayModeAudioTranscription:
 		err = controller.RelayAudioHelper(c, relayMode)
 	default:
-		err = controller.RelayTextHelper(c, relayMode)
+		err = controller.RelayTextHelper(c)
 	}
-	if err != nil {
-		requestId := c.GetString(logger.RequestIdKey)
-		retryTimesStr := c.Query("retry")
-		retryTimes, _ := strconv.Atoi(retryTimesStr)
-		if retryTimesStr == "" {
-			retryTimes = config.RetryTimes
+	return err
+}
+
+func Relay(c *gin.Context) {
+	ctx := c.Request.Context()
+	relayMode := constant.Path2RelayMode(c.Request.URL.Path)
+	if config.DebugEnabled {
+		requestBody, _ := common.GetRequestBody(c)
+		logger.Debugf(ctx, "request body: %s", string(requestBody))
+	}
+	bizErr := relay(c, relayMode)
+	if bizErr == nil {
+		return
+	}
+	channelId := c.GetInt("channel_id")
+	lastFailedChannelId := channelId
+	channelName := c.GetString("channel_name")
+	group := c.GetString("group")
+	originalModel := c.GetString("original_model")
+	go processChannelRelayError(ctx, channelId, channelName, bizErr)
+	requestId := c.GetString(logger.RequestIdKey)
+	retryTimes := config.RetryTimes
+	if !shouldRetry(c, bizErr.StatusCode) {
+		logger.Errorf(ctx, "relay error happen, status code is %d, won't retry in this case", bizErr.StatusCode)
+		retryTimes = 0
+	}
+	for i := retryTimes; i > 0; i-- {
+		channel, err := dbmodel.CacheGetRandomSatisfiedChannel(group, originalModel, i != retryTimes)
+		if err != nil {
+			logger.Errorf(ctx, "CacheGetRandomSatisfiedChannel failed: %w", err)
+			break
 		}
-		if retryTimes > 0 {
-			c.Redirect(http.StatusTemporaryRedirect, fmt.Sprintf("%s?retry=%d", c.Request.URL.Path, retryTimes-1))
-		} else {
-			if err.StatusCode == http.StatusTooManyRequests {
-				err.Error.Message = "当前分组上游负载已饱和，请稍后再试"
-			}
-			err.Error.Message = helper.MessageWithRequestId(err.Error.Message, requestId)
-			c.JSON(err.StatusCode, gin.H{
-				"error": err.Error,
-			})
+		logger.Infof(ctx, "using channel #%d to retry (remain times %d)", channel.Id, i)
+		if channel.Id == lastFailedChannelId {
+			continue
+		}
+		middleware.SetupContextForSelectedChannel(c, channel, originalModel)
+		requestBody, err := common.GetRequestBody(c)
+		c.Request.Body = io.NopCloser(bytes.NewBuffer(requestBody))
+		bizErr = relay(c, relayMode)
+		if bizErr == nil {
+			return
 		}
 		channelId := c.GetInt("channel_id")
-		logger.Error(c.Request.Context(), fmt.Sprintf("relay error (channel #%d): %s", channelId, err.Message))
-		// https://platform.openai.com/docs/guides/error-codes/api-errors
-		if util.ShouldDisableChannel(&err.Error, err.StatusCode) {
-			channelId := c.GetInt("channel_id")
-			channelName := c.GetString("channel_name")
-			disableChannel(channelId, channelName, err.Message)
+		lastFailedChannelId = channelId
+		channelName := c.GetString("channel_name")
+		go processChannelRelayError(ctx, channelId, channelName, bizErr)
+	}
+	if bizErr != nil {
+		if bizErr.StatusCode == http.StatusTooManyRequests {
+			bizErr.Error.Message = "当前分组上游负载已饱和，请稍后再试"
 		}
+		bizErr.Error.Message = helper.MessageWithRequestId(bizErr.Error.Message, requestId)
+		c.JSON(bizErr.StatusCode, gin.H{
+			"error": bizErr.Error,
+		})
+	}
+}
+
+func shouldRetry(c *gin.Context, statusCode int) bool {
+	if _, ok := c.Get("specific_channel_id"); ok {
+		return false
+	}
+	if statusCode == http.StatusTooManyRequests {
+		return true
+	}
+	if statusCode/100 == 5 {
+		return true
+	}
+	if statusCode == http.StatusBadRequest {
+		return false
+	}
+	if statusCode/100 == 2 {
+		return false
+	}
+	return true
+}
+
+func processChannelRelayError(ctx context.Context, channelId int, channelName string, err *model.ErrorWithStatusCode) {
+	logger.Errorf(ctx, "relay error (channel #%d): %s", channelId, err.Message)
+	// https://platform.openai.com/docs/guides/error-codes/api-errors
+	if util.ShouldDisableChannel(&err.Error, err.StatusCode) {
+		disableChannel(channelId, channelName, err.Message)
 	}
 }
 
 func RelayNotImplemented(c *gin.Context) {
-	err := openai.Error{
+	err := model.Error{
 		Message: "API not implemented",
 		Type:    "one_api_error",
 		Param:   "",
@@ -73,7 +134,7 @@ func RelayNotImplemented(c *gin.Context) {
 }
 
 func RelayNotFound(c *gin.Context) {
-	err := openai.Error{
+	err := model.Error{
 		Message: fmt.Sprintf("Invalid URL (%s %s)", c.Request.Method, c.Request.URL.Path),
 		Type:    "invalid_request_error",
 		Param:   "",

controller/token.go

@@ -2,11 +2,11 @@ package controller
 
 import (
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/model"
 	"strconv"
 )
 

controller/user.go

@@ -3,11 +3,11 @@ package controller
 import (
 	"encoding/json"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/model"
 	"strconv"
 	"time"
 

controller/wechat.go

@@ -5,10 +5,10 @@ import (
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/model"
 	"strconv"
 	"time"
 )

go.mod

@@ -1,4 +1,4 @@
-module one-api
+module github.com/songquanpeng/one-api
 
 // +heroku goVersion go1.18
 go 1.18

i18n/en.json

@@ -456,6 +456,7 @@
   "已绑定的邮箱账户": "Email Account Bound",
   "用户信息更新成功！": "User information updated successfully!",
   "模型倍率 %.2f，分组倍率 %.2f": "model rate %.2f, group rate %.2f",
+  "模型倍率 %.2f，分组倍率 %.2f，补全倍率 %.2f": "model rate %.2f, group rate %.2f, completion rate %.2f",
   "使用明细（总消耗额度：{renderQuota(stat.quota)}）": "Usage Details (Total Consumption Quota: {renderQuota(stat.quota)})",
   "用户名称": "User Name",
   "令牌名称": "Token Name",

main.go

@@ -6,14 +6,14 @@ import (
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-contrib/sessions/cookie"
 	"github.com/gin-gonic/gin"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/logger"
-	"one-api/controller"
-	"one-api/middleware"
-	"one-api/model"
-	"one-api/relay/channel/openai"
-	"one-api/router"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/controller"
+	"github.com/songquanpeng/one-api/middleware"
+	"github.com/songquanpeng/one-api/model"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/router"
 	"os"
 	"strconv"
 )

middleware/auth.go

@@ -3,9 +3,9 @@ package middleware
 import (
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/model"
 	"strings"
 )
 
@@ -108,7 +108,7 @@ func TokenAuth() func(c *gin.Context) {
 		c.Set("token_name", token.Name)
 		if len(parts) > 1 {
 			if model.IsAdmin(token.UserId) {
-				c.Set("channelId", parts[1])
+				c.Set("specific_channel_id", parts[1])
 			} else {
 				abortWithMessage(c, http.StatusForbidden, "普通用户不支持指定渠道")
 				return

middleware/distributor.go

@@ -2,10 +2,10 @@ package middleware
 
 import (
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/model"
 	"net/http"
-	"one-api/common"
-	"one-api/common/logger"
-	"one-api/model"
 	"strconv"
 	"strings"
 
@@ -21,8 +21,9 @@ func Distribute() func(c *gin.Context) {
 		userId := c.GetInt("id")
 		userGroup, _ := model.CacheGetUserGroup(userId)
 		c.Set("group", userGroup)
+		var requestModel string
 		var channel *model.Channel
-		channelId, ok := c.Get("channelId")
+		channelId, ok := c.Get("specific_channel_id")
 		if ok {
 			id, err := strconv.Atoi(channelId.(string))
 			if err != nil {
@@ -66,7 +67,8 @@ func Distribute() func(c *gin.Context) {
 					modelRequest.Model = "whisper-1"
 				}
 			}
-			channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, modelRequest.Model)
+			requestModel = modelRequest.Model
+			channel, err = model.CacheGetRandomSatisfiedChannel(userGroup, modelRequest.Model, false)
 			if err != nil {
 				message := fmt.Sprintf("当前分组 %s 下对于模型 %s 无可用渠道", userGroup, modelRequest.Model)
 				if channel != nil {
@@ -77,24 +79,34 @@ func Distribute() func(c *gin.Context) {
 				return
 			}
 		}
-		c.Set("channel", channel.Type)
-		c.Set("channel_id", channel.Id)
-		c.Set("channel_name", channel.Name)
-		c.Set("model_mapping", channel.GetModelMapping())
-		c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
-		c.Set("base_url", channel.GetBaseURL())
-		switch channel.Type {
-		case common.ChannelTypeAzure:
-			c.Set("api_version", channel.Other)
-		case common.ChannelTypeXunfei:
-			c.Set("api_version", channel.Other)
-		case common.ChannelTypeGemini:
-			c.Set("api_version", channel.Other)
-		case common.ChannelTypeAIProxyLibrary:
-			c.Set("library_id", channel.Other)
-		case common.ChannelTypeAli:
-			c.Set("plugin", channel.Other)
-		}
+		SetupContextForSelectedChannel(c, channel, requestModel)
 		c.Next()
 	}
 }
+
+func SetupContextForSelectedChannel(c *gin.Context, channel *model.Channel, modelName string) {
+	c.Set("channel", channel.Type)
+	c.Set("channel_id", channel.Id)
+	c.Set("channel_name", channel.Name)
+	c.Set("model_mapping", channel.GetModelMapping())
+	c.Set("original_model", modelName) // for retry
+	c.Request.Header.Set("Authorization", fmt.Sprintf("Bearer %s", channel.Key))
+	c.Set("base_url", channel.GetBaseURL())
+	// this is for backward compatibility
+	switch channel.Type {
+	case common.ChannelTypeAzure:
+		c.Set(common.ConfigKeyAPIVersion, channel.Other)
+	case common.ChannelTypeXunfei:
+		c.Set(common.ConfigKeyAPIVersion, channel.Other)
+	case common.ChannelTypeGemini:
+		c.Set(common.ConfigKeyAPIVersion, channel.Other)
+	case common.ChannelTypeAIProxyLibrary:
+		c.Set(common.ConfigKeyLibraryID, channel.Other)
+	case common.ChannelTypeAli:
+		c.Set(common.ConfigKeyPlugin, channel.Other)
+	}
+	cfg, _ := channel.LoadConfig()
+	for k, v := range cfg {
+		c.Set(common.ConfigKeyPrefix+k, v)
+	}
+}

middleware/logger.go

@@ -3,7 +3,7 @@ package middleware
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/logger"
 )
 
 func SetUpLogger(server *gin.Engine) {

middleware/rate-limit.go

@@ -4,9 +4,9 @@ import (
 	"context"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
 	"time"
 )
 

middleware/recover.go

@@ -3,8 +3,8 @@ package middleware
 import (
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/logger"
 	"net/http"
-	"one-api/common/logger"
 	"runtime/debug"
 )
 

middleware/request-id.go

@@ -3,13 +3,13 @@ package middleware
 import (
 	"context"
 	"github.com/gin-gonic/gin"
-	"one-api/common/helper"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 )
 
 func RequestId() func(c *gin.Context) {
 	return func(c *gin.Context) {
-		id := helper.GetTimeString() + helper.GetRandomString(8)
+		id := helper.GetTimeString() + helper.GetRandomNumberString(8)
 		c.Set(logger.RequestIdKey, id)
 		ctx := context.WithValue(c.Request.Context(), logger.RequestIdKey, id)
 		c.Request = c.Request.WithContext(ctx)

middleware/turnstile-check.go

@@ -4,10 +4,10 @@ import (
 	"encoding/json"
 	"github.com/gin-contrib/sessions"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
 	"net/http"
 	"net/url"
-	"one-api/common/config"
-	"one-api/common/logger"
 )
 
 type turnstileCheckResponse struct {

middleware/utils.go

@@ -2,8 +2,8 @@ package middleware
 
 import (
 	"github.com/gin-gonic/gin"
-	"one-api/common/helper"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 )
 
 func abortWithMessage(c *gin.Context, statusCode int, message string) {

model/ability.go

@@ -1,7 +1,7 @@
 package model
 
 import (
-	"one-api/common"
+	"github.com/songquanpeng/one-api/common"
 	"strings"
 )
 

model/cache.go

@@ -4,10 +4,10 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
 	"math/rand"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/logger"
 	"sort"
 	"strconv"
 	"strings"
@@ -94,7 +94,7 @@ func CacheUpdateUserQuota(id int) error {
 	if !common.RedisEnabled {
 		return nil
 	}
-	quota, err := GetUserQuota(id)
+	quota, err := CacheGetUserQuota(id)
 	if err != nil {
 		return err
 	}
@@ -191,7 +191,7 @@ func SyncChannelCache(frequency int) {
 	}
 }
 
-func CacheGetRandomSatisfiedChannel(group string, model string) (*Channel, error) {
+func CacheGetRandomSatisfiedChannel(group string, model string, ignoreFirstPriority bool) (*Channel, error) {
 	if !config.MemoryCacheEnabled {
 		return GetRandomSatisfiedChannel(group, model)
 	}
@@ -213,5 +213,10 @@ func CacheGetRandomSatisfiedChannel(group string, model string) (*Channel, error
 		}
 	}
 	idx := rand.Intn(endIdx)
+	if ignoreFirstPriority {
+		if endIdx < len(channels) { // which means there are more than one priority
+			idx = common.RandRange(endIdx, len(channels))
+		}
+	}
 	return channels[idx], nil
 }

model/channel.go

@@ -3,11 +3,11 @@ package model
 import (
 	"encoding/json"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 	"gorm.io/gorm"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
 )
 
 type Channel struct {
@@ -21,7 +21,7 @@ type Channel struct {
 	TestTime           int64   `json:"test_time" gorm:"bigint"`
 	ResponseTime       int     `json:"response_time"` // in milliseconds
 	BaseURL            *string `json:"base_url" gorm:"column:base_url;default:''"`
-	Other              string  `json:"other"`
+	Other              string  `json:"other"`   // DEPRECATED: please save config to field Config
 	Balance            float64 `json:"balance"` // in USD
 	BalanceUpdatedTime int64   `json:"balance_updated_time" gorm:"bigint"`
 	Models             string  `json:"models"`
@@ -29,6 +29,7 @@ type Channel struct {
 	UsedQuota          int64   `json:"used_quota" gorm:"bigint;default:0"`
 	ModelMapping       *string `json:"model_mapping" gorm:"type:varchar(1024);default:''"`
 	Priority           *int64  `json:"priority" gorm:"bigint;default:0"`
+	Config             string  `json:"config"`
 }
 
 func GetAllChannels(startIdx int, num int, selectAll bool) ([]*Channel, error) {
@@ -155,6 +156,18 @@ func (channel *Channel) Delete() error {
 	return err
 }
 
+func (channel *Channel) LoadConfig() (map[string]string, error) {
+	if channel.Config == "" {
+		return nil, nil
+	}
+	cfg := make(map[string]string)
+	err := json.Unmarshal([]byte(channel.Config), &cfg)
+	if err != nil {
+		return nil, err
+	}
+	return cfg, nil
+}
+
 func UpdateChannelStatusById(id int, status int) {
 	err := UpdateAbilityStatus(id, status == common.ChannelStatusEnabled)
 	if err != nil {

model/log.go

@@ -3,10 +3,10 @@ package model
 import (
 	"context"
 	"fmt"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 
 	"gorm.io/gorm"
 )

model/main.go

@@ -2,14 +2,14 @@ package model
 
 import (
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 	"gorm.io/driver/mysql"
 	"gorm.io/driver/postgres"
 	"gorm.io/driver/sqlite"
 	"gorm.io/gorm"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
 	"os"
 	"strings"
 	"time"

model/option.go

@@ -1,9 +1,9 @@
 package model
 
 import (
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
 	"strconv"
 	"strings"
 	"time"
@@ -66,6 +66,7 @@ func InitOptionMap() {
 	config.OptionMap["PreConsumedQuota"] = strconv.Itoa(config.PreConsumedQuota)
 	config.OptionMap["ModelRatio"] = common.ModelRatio2JSONString()
 	config.OptionMap["GroupRatio"] = common.GroupRatio2JSONString()
+	config.OptionMap["CompletionRatio"] = common.CompletionRatio2JSONString()
 	config.OptionMap["TopUpLink"] = config.TopUpLink
 	config.OptionMap["ChatLink"] = config.ChatLink
 	config.OptionMap["QuotaPerUnit"] = strconv.FormatFloat(config.QuotaPerUnit, 'f', -1, 64)
@@ -198,6 +199,8 @@ func updateOptionMap(key string, value string) (err error) {
 		err = common.UpdateModelRatioByJSONString(value)
 	case "GroupRatio":
 		err = common.UpdateGroupRatioByJSONString(value)
+	case "CompletionRatio":
+		err = common.UpdateCompletionRatioByJSONString(value)
 	case "TopUpLink":
 		config.TopUpLink = value
 	case "ChatLink":

model/redemption.go

@@ -3,9 +3,9 @@ package model
 import (
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
 	"gorm.io/gorm"
-	"one-api/common"
-	"one-api/common/helper"
 )
 
 type Redemption struct {

model/token.go

@@ -3,11 +3,11 @@ package model
 import (
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 	"gorm.io/gorm"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
 )
 
 type Token struct {

model/user.go

@@ -3,11 +3,11 @@ package model
 import (
 	"errors"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
 	"gorm.io/gorm"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/logger"
 	"strings"
 )
 

model/utils.go

@@ -1,8 +1,8 @@
 package model
 
 import (
-	"one-api/common/config"
-	"one-api/common/logger"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/logger"
 	"sync"
 	"time"
 )

relay/channel/ai360/constants.go

@@ -0,0 +1,8 @@
+package ai360
+
+var ModelList = []string{
+	"360GPT_S2_V9",
+	"embedding-bert-512-v1",
+	"embedding_s1_v1",
+	"semantic_similarity_s1_v1",
+}

relay/channel/aiproxy/adaptor.go

@@ -1,22 +1,60 @@
 package aiproxy
 
 import (
+	"errors"
+	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/relay/channel/openai"
 )
 
 type Adaptor struct {
 }
 
-func (a *Adaptor) Auth(c *gin.Context) error {
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	return fmt.Sprintf("%s/api/library/ask", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
 	return nil
 }
 
-func (a *Adaptor) ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error) {
-	return nil, nil
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	aiProxyLibraryRequest := ConvertRequest(*request)
+	aiProxyLibraryRequest.LibraryId = c.GetString(common.ConfigKeyLibraryID)
+	return aiProxyLibraryRequest, nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error) {
-	return nil, nil, nil
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err, usage = StreamHandler(c, resp)
+	} else {
+		err, usage = Handler(c, resp)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "aiproxy"
 }

relay/channel/aiproxy/constants.go

@@ -0,0 +1,9 @@
+package aiproxy
+
+import "github.com/songquanpeng/one-api/relay/channel/openai"
+
+var ModelList = []string{""}
+
+func init() {
+	ModelList = openai.ModelList
+}

relay/channel/aiproxy/main.go

@@ -5,20 +5,21 @@ import (
 	"encoding/json"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/helper"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
-	"one-api/relay/constant"
 	"strconv"
 	"strings"
 )
 
 // https://docs.aiproxy.io/dev/library#使用已经定制好的知识库进行对话问答
 
-func ConvertRequest(request openai.GeneralOpenAIRequest) *LibraryRequest {
+func ConvertRequest(request model.GeneralOpenAIRequest) *LibraryRequest {
 	query := ""
 	if len(request.Messages) != 0 {
 		query = request.Messages[len(request.Messages)-1].StringContent()
@@ -45,14 +46,14 @@ func responseAIProxyLibrary2OpenAI(response *LibraryResponse) *openai.TextRespon
 	content := response.Answer + aiProxyDocuments2Markdown(response.Documents)
 	choice := openai.TextResponseChoice{
 		Index: 0,
-		Message: openai.Message{
+		Message: model.Message{
 			Role:    "assistant",
 			Content: content,
 		},
 		FinishReason: "stop",
 	}
 	fullTextResponse := openai.TextResponse{
-		Id:      helper.GetUUID(),
+		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
 		Object:  "chat.completion",
 		Created: helper.GetTimestamp(),
 		Choices: []openai.TextResponseChoice{choice},
@@ -65,7 +66,7 @@ func documentsAIProxyLibrary(documents []LibraryDocument) *openai.ChatCompletion
 	choice.Delta.Content = aiProxyDocuments2Markdown(documents)
 	choice.FinishReason = &constant.StopFinishReason
 	return &openai.ChatCompletionsStreamResponse{
-		Id:      helper.GetUUID(),
+		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
 		Object:  "chat.completion.chunk",
 		Created: helper.GetTimestamp(),
 		Model:   "",
@@ -77,7 +78,7 @@ func streamResponseAIProxyLibrary2OpenAI(response *LibraryStreamResponse) *opena
 	var choice openai.ChatCompletionsStreamResponseChoice
 	choice.Delta.Content = response.Content
 	return &openai.ChatCompletionsStreamResponse{
-		Id:      helper.GetUUID(),
+		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
 		Object:  "chat.completion.chunk",
 		Created: helper.GetTimestamp(),
 		Model:   response.Model,
@@ -85,8 +86,8 @@ func streamResponseAIProxyLibrary2OpenAI(response *LibraryStreamResponse) *opena
 	}
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
-	var usage openai.Usage
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var usage model.Usage
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
 		if atEOF && len(data) == 0 {
@@ -157,7 +158,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatus
 	return nil, &usage
 }
 
-func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var AIProxyLibraryResponse LibraryResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
@@ -172,8 +173,8 @@ func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode,
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if AIProxyLibraryResponse.ErrCode != 0 {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: AIProxyLibraryResponse.Message,
 				Type:    strconv.Itoa(AIProxyLibraryResponse.ErrCode),
 				Code:    AIProxyLibraryResponse.ErrCode,
@@ -189,5 +190,8 @@ func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode,
 	c.Writer.Header().Set("Content-Type", "application/json")
 	c.Writer.WriteHeader(resp.StatusCode)
 	_, err = c.Writer.Write(jsonResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "write_response_body_failed", http.StatusInternalServerError), nil
+	}
 	return nil, &fullTextResponse.Usage
 }

relay/channel/ali/adaptor.go

@@ -1,22 +1,83 @@
 package ali
 
 import (
+	"errors"
+	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/relay/channel/openai"
 )
 
+// https://help.aliyun.com/zh/dashscope/developer-reference/api-details
+
 type Adaptor struct {
 }
 
-func (a *Adaptor) Auth(c *gin.Context) error {
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	fullRequestURL := fmt.Sprintf("%s/api/v1/services/aigc/text-generation/generation", meta.BaseURL)
+	if meta.Mode == constant.RelayModeEmbeddings {
+		fullRequestURL = fmt.Sprintf("%s/api/v1/services/embeddings/text-embedding/text-embedding", meta.BaseURL)
+	}
+	return fullRequestURL, nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
+	if meta.IsStream {
+		req.Header.Set("X-DashScope-SSE", "enable")
+	}
+	if c.GetString(common.ConfigKeyPlugin) != "" {
+		req.Header.Set("X-DashScope-Plugin", c.GetString(common.ConfigKeyPlugin))
+	}
 	return nil
 }
 
-func (a *Adaptor) ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error) {
-	return nil, nil
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	switch relayMode {
+	case constant.RelayModeEmbeddings:
+		baiduEmbeddingRequest := ConvertEmbeddingRequest(*request)
+		return baiduEmbeddingRequest, nil
+	default:
+		baiduRequest := ConvertRequest(*request)
+		return baiduRequest, nil
+	}
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error) {
-	return nil, nil, nil
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err, usage = StreamHandler(c, resp)
+	} else {
+		switch meta.Mode {
+		case constant.RelayModeEmbeddings:
+			err, usage = EmbeddingHandler(c, resp)
+		default:
+			err, usage = Handler(c, resp)
+		}
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "ali"
 }

relay/channel/ali/constants.go

@@ -0,0 +1,6 @@
+package ali
+
+var ModelList = []string{
+	"qwen-turbo", "qwen-plus", "qwen-max", "qwen-max-longcontext",
+	"text-embedding-v1",
+}

relay/channel/ali/main.go

@@ -4,12 +4,13 @@ import (
 	"bufio"
 	"encoding/json"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/helper"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
 	"strings"
 )
 
@@ -17,7 +18,7 @@ import (
 
 const EnableSearchModelSuffix = "-internet"
 
-func ConvertRequest(request openai.GeneralOpenAIRequest) *ChatRequest {
+func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
 	messages := make([]Message, 0, len(request.Messages))
 	for i := 0; i < len(request.Messages); i++ {
 		message := request.Messages[i]
@@ -32,6 +33,9 @@ func ConvertRequest(request openai.GeneralOpenAIRequest) *ChatRequest {
 		enableSearch = true
 		aliModel = strings.TrimSuffix(aliModel, EnableSearchModelSuffix)
 	}
+	if request.TopP >= 1 {
+		request.TopP = 0.9999
+	}
 	return &ChatRequest{
 		Model: aliModel,
 		Input: Input{
@@ -40,11 +44,15 @@ func ConvertRequest(request openai.GeneralOpenAIRequest) *ChatRequest {
 		Parameters: Parameters{
 			EnableSearch:      enableSearch,
 			IncrementalOutput: request.Stream,
+			Seed:              uint64(request.Seed),
+			MaxTokens:         request.MaxTokens,
+			Temperature:       request.Temperature,
+			TopP:              request.TopP,
 		},
 	}
 }
 
-func ConvertEmbeddingRequest(request openai.GeneralOpenAIRequest) *EmbeddingRequest {
+func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *EmbeddingRequest {
 	return &EmbeddingRequest{
 		Model: "text-embedding-v1",
 		Input: struct {
@@ -55,7 +63,7 @@ func ConvertEmbeddingRequest(request openai.GeneralOpenAIRequest) *EmbeddingRequ
 	}
 }
 
-func EmbeddingHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func EmbeddingHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var aliResponse EmbeddingResponse
 	err := json.NewDecoder(resp.Body).Decode(&aliResponse)
 	if err != nil {
@@ -68,8 +76,8 @@ func EmbeddingHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithSta
 	}
 
 	if aliResponse.Code != "" {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: aliResponse.Message,
 				Type:    aliResponse.Code,
 				Param:   aliResponse.RequestId,
@@ -95,7 +103,7 @@ func embeddingResponseAli2OpenAI(response *EmbeddingResponse) *openai.EmbeddingR
 		Object: "list",
 		Data:   make([]openai.EmbeddingResponseItem, 0, len(response.Output.Embeddings)),
 		Model:  "text-embedding-v1",
-		Usage:  openai.Usage{TotalTokens: response.Usage.TotalTokens},
+		Usage:  model.Usage{TotalTokens: response.Usage.TotalTokens},
 	}
 
 	for _, item := range response.Output.Embeddings {
@@ -111,7 +119,7 @@ func embeddingResponseAli2OpenAI(response *EmbeddingResponse) *openai.EmbeddingR
 func responseAli2OpenAI(response *ChatResponse) *openai.TextResponse {
 	choice := openai.TextResponseChoice{
 		Index: 0,
-		Message: openai.Message{
+		Message: model.Message{
 			Role:    "assistant",
 			Content: response.Output.Text,
 		},
@@ -122,7 +130,7 @@ func responseAli2OpenAI(response *ChatResponse) *openai.TextResponse {
 		Object:  "chat.completion",
 		Created: helper.GetTimestamp(),
 		Choices: []openai.TextResponseChoice{choice},
-		Usage: openai.Usage{
+		Usage: model.Usage{
 			PromptTokens:     response.Usage.InputTokens,
 			CompletionTokens: response.Usage.OutputTokens,
 			TotalTokens:      response.Usage.InputTokens + response.Usage.OutputTokens,
@@ -148,8 +156,8 @@ func streamResponseAli2OpenAI(aliResponse *ChatResponse) *openai.ChatCompletions
 	return &response
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
-	var usage openai.Usage
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var usage model.Usage
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
 		if atEOF && len(data) == 0 {
@@ -217,7 +225,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatus
 	return nil, &usage
 }
 
-func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var aliResponse ChatResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
@@ -232,8 +240,8 @@ func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode,
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if aliResponse.Code != "" {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: aliResponse.Message,
 				Type:    aliResponse.Code,
 				Param:   aliResponse.RequestId,

relay/channel/ali/model.go

@@ -16,6 +16,8 @@ type Parameters struct {
 	Seed              uint64  `json:"seed,omitempty"`
 	EnableSearch      bool    `json:"enable_search,omitempty"`
 	IncrementalOutput bool    `json:"incremental_output,omitempty"`
+	MaxTokens         int     `json:"max_tokens,omitempty"`
+	Temperature       float64 `json:"temperature,omitempty"`
 }
 
 type ChatRequest struct {

relay/channel/anthropic/adaptor.go

@@ -1,22 +1,65 @@
 package anthropic
 
 import (
+	"errors"
+	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/relay/channel/openai"
 )
 
 type Adaptor struct {
 }
 
-func (a *Adaptor) Auth(c *gin.Context) error {
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	return fmt.Sprintf("%s/v1/complete", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("x-api-key", meta.APIKey)
+	anthropicVersion := c.Request.Header.Get("anthropic-version")
+	if anthropicVersion == "" {
+		anthropicVersion = "2023-06-01"
+	}
+	req.Header.Set("anthropic-version", anthropicVersion)
 	return nil
 }
 
-func (a *Adaptor) ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error) {
-	return nil, nil
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return ConvertRequest(*request), nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error) {
-	return nil, nil, nil
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		var responseText string
+		err, responseText = StreamHandler(c, resp)
+		usage = openai.ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+	} else {
+		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "authropic"
 }

relay/channel/anthropic/constants.go

@@ -0,0 +1,5 @@
+package anthropic
+
+var ModelList = []string{
+	"claude-instant-1", "claude-2", "claude-2.0", "claude-2.1",
+}

relay/channel/anthropic/main.go

@@ -5,12 +5,13 @@ import (
 	"encoding/json"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/helper"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
 	"strings"
 )
 
@@ -25,7 +26,7 @@ func stopReasonClaude2OpenAI(reason string) string {
 	}
 }
 
-func ConvertRequest(textRequest openai.GeneralOpenAIRequest) *Request {
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *Request {
 	claudeRequest := Request{
 		Model:             textRequest.Model,
 		Prompt:            "",
@@ -72,7 +73,7 @@ func streamResponseClaude2OpenAI(claudeResponse *Response) *openai.ChatCompletio
 func responseClaude2OpenAI(claudeResponse *Response) *openai.TextResponse {
 	choice := openai.TextResponseChoice{
 		Index: 0,
-		Message: openai.Message{
+		Message: model.Message{
 			Role:    "assistant",
 			Content: strings.TrimPrefix(claudeResponse.Completion, " "),
 			Name:    nil,
@@ -88,7 +89,7 @@ func responseClaude2OpenAI(claudeResponse *Response) *openai.TextResponse {
 	return &fullTextResponse
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, string) {
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, string) {
 	responseText := ""
 	responseId := fmt.Sprintf("chatcmpl-%s", helper.GetUUID())
 	createdTime := helper.GetTimestamp()
@@ -153,7 +154,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatus
 	return nil, responseText
 }
 
-func Handler(c *gin.Context, resp *http.Response, promptTokens int, model string) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
 		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
@@ -168,8 +169,8 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, model string
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if claudeResponse.Error.Type != "" {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: claudeResponse.Error.Message,
 				Type:    claudeResponse.Error.Type,
 				Param:   "",
@@ -179,9 +180,9 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, model string
 		}, nil
 	}
 	fullTextResponse := responseClaude2OpenAI(&claudeResponse)
-	fullTextResponse.Model = model
-	completionTokens := openai.CountTokenText(claudeResponse.Completion, model)
-	usage := openai.Usage{
+	fullTextResponse.Model = modelName
+	completionTokens := openai.CountTokenText(claudeResponse.Completion, modelName)
+	usage := model.Usage{
 		PromptTokens:     promptTokens,
 		CompletionTokens: completionTokens,
 		TotalTokens:      promptTokens + completionTokens,

relay/channel/baichuan/constants.go

@@ -0,0 +1,7 @@
+package baichuan
+
+var ModelList = []string{
+	"Baichuan2-Turbo",
+	"Baichuan2-Turbo-192k",
+	"Baichuan-Text-Embedding",
+}

relay/channel/baidu/adaptor.go

@@ -1,22 +1,95 @@
 package baidu
 
 import (
+	"errors"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/relay/channel/openai"
 )
 
 type Adaptor struct {
 }
 
-func (a *Adaptor) Auth(c *gin.Context) error {
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	// https://cloud.baidu.com/doc/WENXINWORKSHOP/s/clntwmv7t
+	var fullRequestURL string
+	switch meta.ActualModelName {
+	case "ERNIE-Bot-4":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro"
+	case "ERNIE-Bot-8K":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie_bot_8k"
+	case "ERNIE-Bot":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions"
+	case "ERNIE-Speed":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie_speed"
+	case "ERNIE-Bot-turbo":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/eb-instant"
+	case "BLOOMZ-7B":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/bloomz_7b1"
+	case "Embedding-V1":
+		fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/embeddings/embedding-v1"
+	default:
+               fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/" + meta.ActualModelName
+	}
+	var accessToken string
+	var err error
+	if accessToken, err = GetAccessToken(meta.APIKey); err != nil {
+		return "", err
+	}
+	fullRequestURL += "?access_token=" + accessToken
+	return fullRequestURL, nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
 	return nil
 }
 
-func (a *Adaptor) ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error) {
-	return nil, nil
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	switch relayMode {
+	case constant.RelayModeEmbeddings:
+		baiduEmbeddingRequest := ConvertEmbeddingRequest(*request)
+		return baiduEmbeddingRequest, nil
+	default:
+		baiduRequest := ConvertRequest(*request)
+		return baiduRequest, nil
+	}
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error) {
-	return nil, nil, nil
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		err, usage = StreamHandler(c, resp)
+	} else {
+		switch meta.Mode {
+		case constant.RelayModeEmbeddings:
+			err, usage = EmbeddingHandler(c, resp)
+		default:
+			err, usage = Handler(c, resp)
+		}
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "baidu"
 }

relay/channel/baidu/constants.go

@@ -0,0 +1,10 @@
+package baidu
+
+var ModelList = []string{
+	"ERNIE-Bot-4",
+	"ERNIE-Bot-8K",
+	"ERNIE-Bot",
+	"ERNIE-Speed",
+	"ERNIE-Bot-turbo",
+	"Embedding-V1",
+}

relay/channel/baidu/main.go

@@ -6,13 +6,14 @@ import (
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
-	"one-api/relay/constant"
-	"one-api/relay/util"
 	"strings"
 	"sync"
 	"time"
@@ -43,7 +44,7 @@ type Error struct {
 
 var baiduTokenStore sync.Map
 
-func ConvertRequest(request openai.GeneralOpenAIRequest) *ChatRequest {
+func ConvertRequest(request model.GeneralOpenAIRequest) *ChatRequest {
 	messages := make([]Message, 0, len(request.Messages))
 	for _, message := range request.Messages {
 		if message.Role == "system" {
@@ -71,7 +72,7 @@ func ConvertRequest(request openai.GeneralOpenAIRequest) *ChatRequest {
 func responseBaidu2OpenAI(response *ChatResponse) *openai.TextResponse {
 	choice := openai.TextResponseChoice{
 		Index: 0,
-		Message: openai.Message{
+		Message: model.Message{
 			Role:    "assistant",
 			Content: response.Result,
 		},
@@ -103,7 +104,7 @@ func streamResponseBaidu2OpenAI(baiduResponse *ChatStreamResponse) *openai.ChatC
 	return &response
 }
 
-func ConvertEmbeddingRequest(request openai.GeneralOpenAIRequest) *EmbeddingRequest {
+func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *EmbeddingRequest {
 	return &EmbeddingRequest{
 		Input: request.ParseInput(),
 	}
@@ -126,8 +127,8 @@ func embeddingResponseBaidu2OpenAI(response *EmbeddingResponse) *openai.Embeddin
 	return &openAIEmbeddingResponse
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
-	var usage openai.Usage
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	var usage model.Usage
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
 		if atEOF && len(data) == 0 {
@@ -189,7 +190,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatus
 	return nil, &usage
 }
 
-func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var baiduResponse ChatResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
@@ -204,8 +205,8 @@ func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode,
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if baiduResponse.ErrorMsg != "" {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: baiduResponse.ErrorMsg,
 				Type:    "baidu_error",
 				Param:   "",
@@ -226,7 +227,7 @@ func Handler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode,
 	return nil, &fullTextResponse.Usage
 }
 
-func EmbeddingHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func EmbeddingHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var baiduResponse EmbeddingResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
@@ -241,8 +242,8 @@ func EmbeddingHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithSta
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if baiduResponse.ErrorMsg != "" {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: baiduResponse.ErrorMsg,
 				Type:    "baidu_error",
 				Param:   "",

relay/channel/baidu/model.go

@@ -1,18 +1,18 @@
 package baidu
 
 import (
-	"one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
 	"time"
 )
 
 type ChatResponse struct {
-	Id               string       `json:"id"`
-	Object           string       `json:"object"`
-	Created          int64        `json:"created"`
-	Result           string       `json:"result"`
-	IsTruncated      bool         `json:"is_truncated"`
-	NeedClearHistory bool         `json:"need_clear_history"`
-	Usage            openai.Usage `json:"usage"`
+	Id               string      `json:"id"`
+	Object           string      `json:"object"`
+	Created          int64       `json:"created"`
+	Result           string      `json:"result"`
+	IsTruncated      bool        `json:"is_truncated"`
+	NeedClearHistory bool        `json:"need_clear_history"`
+	Usage            model.Usage `json:"usage"`
 	Error
 }
 
@@ -37,7 +37,7 @@ type EmbeddingResponse struct {
 	Object  string          `json:"object"`
 	Created int64           `json:"created"`
 	Data    []EmbeddingData `json:"data"`
-	Usage   openai.Usage    `json:"usage"`
+	Usage   model.Usage     `json:"usage"`
 	Error
 }
 

relay/channel/common.go

@@ -0,0 +1,51 @@
+package channel
+
+import (
+	"errors"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
+	"net/http"
+)
+
+func SetupCommonRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) {
+	req.Header.Set("Content-Type", c.Request.Header.Get("Content-Type"))
+	req.Header.Set("Accept", c.Request.Header.Get("Accept"))
+	if meta.IsStream && c.Request.Header.Get("Accept") == "" {
+		req.Header.Set("Accept", "text/event-stream")
+	}
+}
+
+func DoRequestHelper(a Adaptor, c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	fullRequestURL, err := a.GetRequestURL(meta)
+	if err != nil {
+		return nil, fmt.Errorf("get request url failed: %w", err)
+	}
+	req, err := http.NewRequest(c.Request.Method, fullRequestURL, requestBody)
+	if err != nil {
+		return nil, fmt.Errorf("new request failed: %w", err)
+	}
+	err = a.SetupRequestHeader(c, req, meta)
+	if err != nil {
+		return nil, fmt.Errorf("setup request header failed: %w", err)
+	}
+	resp, err := DoRequest(c, req)
+	if err != nil {
+		return nil, fmt.Errorf("do request failed: %w", err)
+	}
+	return resp, nil
+}
+
+func DoRequest(c *gin.Context, req *http.Request) (*http.Response, error) {
+	resp, err := util.HTTPClient.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	if resp == nil {
+		return nil, errors.New("resp is nil")
+	}
+	_ = req.Body.Close()
+	_ = c.Request.Body.Close()
+	return resp, nil
+}

relay/channel/gemini/adaptor.go

@@ -0,0 +1,66 @@
+package gemini
+
+import (
+	"errors"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common/helper"
+	channelhelper "github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
+	"net/http"
+)
+
+type Adaptor struct {
+}
+
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	version := helper.AssignOrDefault(meta.APIVersion, "v1")
+	action := "generateContent"
+	if meta.IsStream {
+		action = "streamGenerateContent"
+	}
+	return fmt.Sprintf("%s/%s/models/%s:%s", meta.BaseURL, version, meta.ActualModelName, action), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channelhelper.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("x-goog-api-key", meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return ConvertRequest(*request), nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channelhelper.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		var responseText string
+		err, responseText = StreamHandler(c, resp)
+		usage = openai.ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+	} else {
+		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "google gemini"
+}

relay/channel/gemini/constants.go

@@ -0,0 +1,6 @@
+package gemini
+
+var ModelList = []string{
+	"gemini-pro", "gemini-1.0-pro-001",
+	"gemini-pro-vision", "gemini-1.0-pro-vision-001",
+}

relay/channel/gemini/main.go

@@ -1,18 +1,19 @@
-package google
+package gemini
 
 import (
 	"bufio"
 	"encoding/json"
 	"fmt"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/image"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/helper"
-	"one-api/common/image"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
-	"one-api/relay/constant"
 	"strings"
 
 	"github.com/gin-gonic/gin"
@@ -21,14 +22,14 @@ import (
 // https://ai.google.dev/docs/gemini_api_overview?hl=zh-cn
 
 const (
-	GeminiVisionMaxImageNum = 16
+	VisionMaxImageNum = 16
 )
 
 // Setting safety to the lowest possible values since Gemini is already powerless enough
-func ConvertGeminiRequest(textRequest openai.GeneralOpenAIRequest) *GeminiChatRequest {
-	geminiRequest := GeminiChatRequest{
-		Contents: make([]GeminiChatContent, 0, len(textRequest.Messages)),
-		SafetySettings: []GeminiChatSafetySettings{
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *ChatRequest {
+	geminiRequest := ChatRequest{
+		Contents: make([]ChatContent, 0, len(textRequest.Messages)),
+		SafetySettings: []ChatSafetySettings{
 			{
 				Category:  "HARM_CATEGORY_HARASSMENT",
 				Threshold: config.GeminiSafetySetting,
@@ -46,14 +47,14 @@ func ConvertGeminiRequest(textRequest openai.GeneralOpenAIRequest) *GeminiChatRe
 				Threshold: config.GeminiSafetySetting,
 			},
 		},
-		GenerationConfig: GeminiChatGenerationConfig{
+		GenerationConfig: ChatGenerationConfig{
 			Temperature:     textRequest.Temperature,
 			TopP:            textRequest.TopP,
 			MaxOutputTokens: textRequest.MaxTokens,
 		},
 	}
 	if textRequest.Functions != nil {
-		geminiRequest.Tools = []GeminiChatTools{
+		geminiRequest.Tools = []ChatTools{
 			{
 				FunctionDeclarations: textRequest.Functions,
 			},
@@ -61,30 +62,30 @@ func ConvertGeminiRequest(textRequest openai.GeneralOpenAIRequest) *GeminiChatRe
 	}
 	shouldAddDummyModelMessage := false
 	for _, message := range textRequest.Messages {
-		content := GeminiChatContent{
+		content := ChatContent{
 			Role: message.Role,
-			Parts: []GeminiPart{
+			Parts: []Part{
 				{
 					Text: message.StringContent(),
 				},
 			},
 		}
 		openaiContent := message.ParseContent()
-		var parts []GeminiPart
+		var parts []Part
 		imageNum := 0
 		for _, part := range openaiContent {
-			if part.Type == openai.ContentTypeText {
-				parts = append(parts, GeminiPart{
+			if part.Type == model.ContentTypeText {
+				parts = append(parts, Part{
 					Text: part.Text,
 				})
-			} else if part.Type == openai.ContentTypeImageURL {
+			} else if part.Type == model.ContentTypeImageURL {
 				imageNum += 1
-				if imageNum > GeminiVisionMaxImageNum {
+				if imageNum > VisionMaxImageNum {
 					continue
 				}
 				mimeType, data, _ := image.GetImageFromUrl(part.ImageURL.Url)
-				parts = append(parts, GeminiPart{
-					InlineData: &GeminiInlineData{
+				parts = append(parts, Part{
+					InlineData: &InlineData{
 						MimeType: mimeType,
 						Data:     data,
 					},
@@ -106,9 +107,9 @@ func ConvertGeminiRequest(textRequest openai.GeneralOpenAIRequest) *GeminiChatRe
 
 		// If a system message is the last message, we need to add a dummy model message to make gemini happy
 		if shouldAddDummyModelMessage {
-			geminiRequest.Contents = append(geminiRequest.Contents, GeminiChatContent{
+			geminiRequest.Contents = append(geminiRequest.Contents, ChatContent{
 				Role: "model",
-				Parts: []GeminiPart{
+				Parts: []Part{
 					{
 						Text: "Okay",
 					},
@@ -121,12 +122,12 @@ func ConvertGeminiRequest(textRequest openai.GeneralOpenAIRequest) *GeminiChatRe
 	return &geminiRequest
 }
 
-type GeminiChatResponse struct {
-	Candidates     []GeminiChatCandidate    `json:"candidates"`
-	PromptFeedback GeminiChatPromptFeedback `json:"promptFeedback"`
+type ChatResponse struct {
+	Candidates     []ChatCandidate    `json:"candidates"`
+	PromptFeedback ChatPromptFeedback `json:"promptFeedback"`
 }
 
-func (g *GeminiChatResponse) GetResponseText() string {
+func (g *ChatResponse) GetResponseText() string {
 	if g == nil {
 		return ""
 	}
@@ -136,23 +137,23 @@ func (g *GeminiChatResponse) GetResponseText() string {
 	return ""
 }
 
-type GeminiChatCandidate struct {
-	Content       GeminiChatContent        `json:"content"`
-	FinishReason  string                   `json:"finishReason"`
-	Index         int64                    `json:"index"`
-	SafetyRatings []GeminiChatSafetyRating `json:"safetyRatings"`
+type ChatCandidate struct {
+	Content       ChatContent        `json:"content"`
+	FinishReason  string             `json:"finishReason"`
+	Index         int64              `json:"index"`
+	SafetyRatings []ChatSafetyRating `json:"safetyRatings"`
 }
 
-type GeminiChatSafetyRating struct {
+type ChatSafetyRating struct {
 	Category    string `json:"category"`
 	Probability string `json:"probability"`
 }
 
-type GeminiChatPromptFeedback struct {
-	SafetyRatings []GeminiChatSafetyRating `json:"safetyRatings"`
+type ChatPromptFeedback struct {
+	SafetyRatings []ChatSafetyRating `json:"safetyRatings"`
 }
 
-func responseGeminiChat2OpenAI(response *GeminiChatResponse) *openai.TextResponse {
+func responseGeminiChat2OpenAI(response *ChatResponse) *openai.TextResponse {
 	fullTextResponse := openai.TextResponse{
 		Id:      fmt.Sprintf("chatcmpl-%s", helper.GetUUID()),
 		Object:  "chat.completion",
@@ -162,7 +163,7 @@ func responseGeminiChat2OpenAI(response *GeminiChatResponse) *openai.TextRespons
 	for i, candidate := range response.Candidates {
 		choice := openai.TextResponseChoice{
 			Index: i,
-			Message: openai.Message{
+			Message: model.Message{
 				Role:    "assistant",
 				Content: "",
 			},
@@ -176,7 +177,7 @@ func responseGeminiChat2OpenAI(response *GeminiChatResponse) *openai.TextRespons
 	return &fullTextResponse
 }
 
-func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) *openai.ChatCompletionsStreamResponse {
+func streamResponseGeminiChat2OpenAI(geminiResponse *ChatResponse) *openai.ChatCompletionsStreamResponse {
 	var choice openai.ChatCompletionsStreamResponseChoice
 	choice.Delta.Content = geminiResponse.GetResponseText()
 	choice.FinishReason = &constant.StopFinishReason
@@ -187,7 +188,7 @@ func streamResponseGeminiChat2OpenAI(geminiResponse *GeminiChatResponse) *openai
 	return &response
 }
 
-func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, string) {
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, string) {
 	responseText := ""
 	dataChan := make(chan string)
 	stopChan := make(chan bool)
@@ -257,7 +258,7 @@ func StreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatus
 	return nil, responseText
 }
 
-func GeminiHandler(c *gin.Context, resp *http.Response, promptTokens int, model string) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
 		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
@@ -266,14 +267,14 @@ func GeminiHandler(c *gin.Context, resp *http.Response, promptTokens int, model
 	if err != nil {
 		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
-	var geminiResponse GeminiChatResponse
+	var geminiResponse ChatResponse
 	err = json.Unmarshal(responseBody, &geminiResponse)
 	if err != nil {
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if len(geminiResponse.Candidates) == 0 {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: "No candidates returned",
 				Type:    "server_error",
 				Param:   "",
@@ -283,9 +284,9 @@ func GeminiHandler(c *gin.Context, resp *http.Response, promptTokens int, model
 		}, nil
 	}
 	fullTextResponse := responseGeminiChat2OpenAI(&geminiResponse)
-	fullTextResponse.Model = model
-	completionTokens := openai.CountTokenText(geminiResponse.GetResponseText(), model)
-	usage := openai.Usage{
+	fullTextResponse.Model = modelName
+	completionTokens := openai.CountTokenText(geminiResponse.GetResponseText(), modelName)
+	usage := model.Usage{
 		PromptTokens:     promptTokens,
 		CompletionTokens: completionTokens,
 		TotalTokens:      promptTokens + completionTokens,

relay/channel/gemini/model.go

@@ -0,0 +1,41 @@
+package gemini
+
+type ChatRequest struct {
+	Contents         []ChatContent        `json:"contents"`
+	SafetySettings   []ChatSafetySettings `json:"safety_settings,omitempty"`
+	GenerationConfig ChatGenerationConfig `json:"generation_config,omitempty"`
+	Tools            []ChatTools          `json:"tools,omitempty"`
+}
+
+type InlineData struct {
+	MimeType string `json:"mimeType"`
+	Data     string `json:"data"`
+}
+
+type Part struct {
+	Text       string      `json:"text,omitempty"`
+	InlineData *InlineData `json:"inlineData,omitempty"`
+}
+
+type ChatContent struct {
+	Role  string `json:"role,omitempty"`
+	Parts []Part `json:"parts"`
+}
+
+type ChatSafetySettings struct {
+	Category  string `json:"category"`
+	Threshold string `json:"threshold"`
+}
+
+type ChatTools struct {
+	FunctionDeclarations any `json:"functionDeclarations,omitempty"`
+}
+
+type ChatGenerationConfig struct {
+	Temperature     float64  `json:"temperature,omitempty"`
+	TopP            float64  `json:"topP,omitempty"`
+	TopK            float64  `json:"topK,omitempty"`
+	MaxOutputTokens int      `json:"maxOutputTokens,omitempty"`
+	CandidateCount  int      `json:"candidateCount,omitempty"`
+	StopSequences   []string `json:"stopSequences,omitempty"`
+}

relay/channel/google/adaptor.go

@@ -1,22 +0,0 @@
-package google
-
-import (
-	"github.com/gin-gonic/gin"
-	"net/http"
-	"one-api/relay/channel/openai"
-)
-
-type Adaptor struct {
-}
-
-func (a *Adaptor) Auth(c *gin.Context) error {
-	return nil
-}
-
-func (a *Adaptor) ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error) {
-	return nil, nil
-}
-
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error) {
-	return nil, nil, nil
-}

relay/channel/google/model.go

@@ -1,80 +0,0 @@
-package google
-
-import (
-	"one-api/relay/channel/openai"
-)
-
-type GeminiChatRequest struct {
-	Contents         []GeminiChatContent        `json:"contents"`
-	SafetySettings   []GeminiChatSafetySettings `json:"safety_settings,omitempty"`
-	GenerationConfig GeminiChatGenerationConfig `json:"generation_config,omitempty"`
-	Tools            []GeminiChatTools          `json:"tools,omitempty"`
-}
-
-type GeminiInlineData struct {
-	MimeType string `json:"mimeType"`
-	Data     string `json:"data"`
-}
-
-type GeminiPart struct {
-	Text       string            `json:"text,omitempty"`
-	InlineData *GeminiInlineData `json:"inlineData,omitempty"`
-}
-
-type GeminiChatContent struct {
-	Role  string       `json:"role,omitempty"`
-	Parts []GeminiPart `json:"parts"`
-}
-
-type GeminiChatSafetySettings struct {
-	Category  string `json:"category"`
-	Threshold string `json:"threshold"`
-}
-
-type GeminiChatTools struct {
-	FunctionDeclarations any `json:"functionDeclarations,omitempty"`
-}
-
-type GeminiChatGenerationConfig struct {
-	Temperature     float64  `json:"temperature,omitempty"`
-	TopP            float64  `json:"topP,omitempty"`
-	TopK            float64  `json:"topK,omitempty"`
-	MaxOutputTokens int      `json:"maxOutputTokens,omitempty"`
-	CandidateCount  int      `json:"candidateCount,omitempty"`
-	StopSequences   []string `json:"stopSequences,omitempty"`
-}
-
-type PaLMChatMessage struct {
-	Author  string `json:"author"`
-	Content string `json:"content"`
-}
-
-type PaLMFilter struct {
-	Reason  string `json:"reason"`
-	Message string `json:"message"`
-}
-
-type PaLMPrompt struct {
-	Messages []PaLMChatMessage `json:"messages"`
-}
-
-type PaLMChatRequest struct {
-	Prompt         PaLMPrompt `json:"prompt"`
-	Temperature    float64    `json:"temperature,omitempty"`
-	CandidateCount int        `json:"candidateCount,omitempty"`
-	TopP           float64    `json:"topP,omitempty"`
-	TopK           int        `json:"topK,omitempty"`
-}
-
-type PaLMError struct {
-	Code    int    `json:"code"`
-	Message string `json:"message"`
-	Status  string `json:"status"`
-}
-
-type PaLMChatResponse struct {
-	Candidates []PaLMChatMessage `json:"candidates"`
-	Messages   []openai.Message  `json:"messages"`
-	Filters    []PaLMFilter      `json:"filters"`
-	Error      PaLMError         `json:"error"`
-}

relay/channel/interface.go

@@ -2,14 +2,19 @@ package channel
 
 import (
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/relay/channel/openai"
 )
 
 type Adaptor interface {
-	GetRequestURL() string
-	Auth(c *gin.Context) error
-	ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error)
-	DoRequest(request *openai.GeneralOpenAIRequest) error
-	DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error)
+	Init(meta *util.RelayMeta)
+	GetRequestURL(meta *util.RelayMeta) (string, error)
+	SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error
+	ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error)
+	DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error)
+	DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode)
+	GetModelList() []string
+	GetChannelName() string
 }

relay/channel/minimax/constants.go

@@ -0,0 +1,7 @@
+package minimax
+
+var ModelList = []string{
+	"abab5.5s-chat",
+	"abab5.5-chat",
+	"abab6-chat",
+}

relay/channel/minimax/main.go

@@ -0,0 +1,14 @@
+package minimax
+
+import (
+	"fmt"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/util"
+)
+
+func GetRequestURL(meta *util.RelayMeta) (string, error) {
+	if meta.Mode == constant.RelayModeChatCompletions {
+		return fmt.Sprintf("%s/v1/text/chatcompletion_v2", meta.BaseURL), nil
+	}
+	return "", fmt.Errorf("unsupported relay mode %d for minimax", meta.Mode)
+}

relay/channel/mistral/constants.go

@@ -0,0 +1,10 @@
+package mistral
+
+var ModelList = []string{
+	"open-mistral-7b",
+	"open-mixtral-8x7b",
+	"mistral-small-latest",
+	"mistral-medium-latest",
+	"mistral-large-latest",
+	"mistral-embed",
+}

relay/channel/moonshot/constants.go

@@ -0,0 +1,7 @@
+package moonshot
+
+var ModelList = []string{
+	"moonshot-v1-8k",
+	"moonshot-v1-32k",
+	"moonshot-v1-128k",
+}

relay/channel/openai/adaptor.go

@@ -1,21 +1,122 @@
 package openai
 
 import (
+	"errors"
+	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/channel/ai360"
+	"github.com/songquanpeng/one-api/relay/channel/baichuan"
+	"github.com/songquanpeng/one-api/relay/channel/minimax"
+	"github.com/songquanpeng/one-api/relay/channel/mistral"
+	"github.com/songquanpeng/one-api/relay/channel/moonshot"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
+	"strings"
 )
 
 type Adaptor struct {
+	ChannelType int
 }
 
-func (a *Adaptor) Auth(c *gin.Context) error {
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+	a.ChannelType = meta.ChannelType
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	switch meta.ChannelType {
+	case common.ChannelTypeAzure:
+		// https://learn.microsoft.com/en-us/azure/cognitive-services/openai/chatgpt-quickstart?pivots=rest-api&tabs=command-line#rest-api
+		requestURL := strings.Split(meta.RequestURLPath, "?")[0]
+		requestURL = fmt.Sprintf("%s?api-version=%s", requestURL, meta.APIVersion)
+		task := strings.TrimPrefix(requestURL, "/v1/")
+		model_ := meta.ActualModelName
+		model_ = strings.Replace(model_, ".", "", -1)
+		// https://github.com/songquanpeng/one-api/issues/67
+		model_ = strings.TrimSuffix(model_, "-0301")
+		model_ = strings.TrimSuffix(model_, "-0314")
+		model_ = strings.TrimSuffix(model_, "-0613")
+
+		requestURL = fmt.Sprintf("/openai/deployments/%s/%s", model_, task)
+		return util.GetFullRequestURL(meta.BaseURL, requestURL, meta.ChannelType), nil
+	case common.ChannelTypeMinimax:
+		return minimax.GetRequestURL(meta)
+	default:
+		return util.GetFullRequestURL(meta.BaseURL, meta.RequestURLPath, meta.ChannelType), nil
+	}
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	if meta.ChannelType == common.ChannelTypeAzure {
+		req.Header.Set("api-key", meta.APIKey)
+		return nil
+	}
+	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
+	if meta.ChannelType == common.ChannelTypeOpenRouter {
+		req.Header.Set("HTTP-Referer", "https://github.com/songquanpeng/one-api")
+		req.Header.Set("X-Title", "One API")
+	}
 	return nil
 }
 
-func (a *Adaptor) ConvertRequest(request *GeneralOpenAIRequest) (any, error) {
-	return nil, nil
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return request, nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*ErrorWithStatusCode, *Usage, error) {
-	return nil, nil, nil
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		var responseText string
+		err, responseText, _ = StreamHandler(c, resp, meta.Mode)
+		usage = ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+	} else {
+		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	switch a.ChannelType {
+	case common.ChannelType360:
+		return ai360.ModelList
+	case common.ChannelTypeMoonshot:
+		return moonshot.ModelList
+	case common.ChannelTypeBaichuan:
+		return baichuan.ModelList
+	case common.ChannelTypeMinimax:
+		return minimax.ModelList
+	case common.ChannelTypeMistral:
+		return mistral.ModelList
+	default:
+		return ModelList
+	}
+}
+
+func (a *Adaptor) GetChannelName() string {
+	switch a.ChannelType {
+	case common.ChannelTypeAzure:
+		return "azure"
+	case common.ChannelType360:
+		return "360"
+	case common.ChannelTypeMoonshot:
+		return "moonshot"
+	case common.ChannelTypeBaichuan:
+		return "baichuan"
+	case common.ChannelTypeMinimax:
+		return "minimax"
+	case common.ChannelTypeMistral:
+		return "mistralai"
+	default:
+		return "openai"
+	}
 }

relay/channel/openai/constants.go

@@ -0,0 +1,19 @@
+package openai
+
+var ModelList = []string{
+	"gpt-3.5-turbo", "gpt-3.5-turbo-0301", "gpt-3.5-turbo-0613", "gpt-3.5-turbo-1106", "gpt-3.5-turbo-0125",
+	"gpt-3.5-turbo-16k", "gpt-3.5-turbo-16k-0613",
+	"gpt-3.5-turbo-instruct",
+	"gpt-4", "gpt-4-0314", "gpt-4-0613", "gpt-4-1106-preview", "gpt-4-0125-preview",
+	"gpt-4-32k", "gpt-4-32k-0314", "gpt-4-32k-0613",
+	"gpt-4-turbo-preview",
+	"gpt-4-vision-preview",
+	"text-embedding-ada-002", "text-embedding-3-small", "text-embedding-3-large",
+	"text-curie-001", "text-babbage-001", "text-ada-001", "text-davinci-002", "text-davinci-003",
+	"text-moderation-latest", "text-moderation-stable",
+	"text-davinci-edit-001",
+	"davinci-002", "babbage-002",
+	"dall-e-2", "dall-e-3",
+	"whisper-1",
+	"tts-1", "tts-1-1106", "tts-1-hd", "tts-1-hd-1106",
+}

relay/channel/openai/helper.go

@@ -0,0 +1,11 @@
+package openai
+
+import "github.com/songquanpeng/one-api/relay/model"
+
+func ResponseText2Usage(responseText string, modeName string, promptTokens int) *model.Usage {
+	usage := &model.Usage{}
+	usage.PromptTokens = promptTokens
+	usage.CompletionTokens = CountTokenText(responseText, modeName)
+	usage.TotalTokens = usage.PromptTokens + usage.CompletionTokens
+	return usage
+}

relay/channel/openai/main.go

@@ -5,15 +5,16 @@ import (
 	"bytes"
 	"encoding/json"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/logger"
-	"one-api/relay/constant"
 	"strings"
 )
 
-func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*ErrorWithStatusCode, string) {
+func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*model.ErrorWithStatusCode, string, *model.Usage) {
 	responseText := ""
 	scanner := bufio.NewScanner(resp.Body)
 	scanner.Split(func(data []byte, atEOF bool) (advance int, token []byte, err error) {
@@ -30,6 +31,7 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*ErrorWi
 	})
 	dataChan := make(chan string)
 	stopChan := make(chan bool)
+	var usage *model.Usage
 	go func() {
 		for scanner.Scan() {
 			data := scanner.Text()
@@ -53,6 +55,9 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*ErrorWi
 					for _, choice := range streamResponse.Choices {
 						responseText += choice.Delta.Content
 					}
+					if streamResponse.Usage != nil {
+						usage = streamResponse.Usage
+					}
 				case constant.RelayModeCompletions:
 					var streamResponse CompletionsStreamResponse
 					err := json.Unmarshal([]byte(data), &streamResponse)
@@ -85,12 +90,12 @@ func StreamHandler(c *gin.Context, resp *http.Response, relayMode int) (*ErrorWi
 	})
 	err := resp.Body.Close()
 	if err != nil {
-		return ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), ""
+		return ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), "", nil
 	}
-	return nil, responseText
+	return nil, responseText, usage
 }
 
-func Handler(c *gin.Context, resp *http.Response, promptTokens int, model string) (*ErrorWithStatusCode, *Usage) {
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
 	var textResponse SlimTextResponse
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
@@ -105,7 +110,7 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, model string
 		return ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if textResponse.Error.Type != "" {
-		return &ErrorWithStatusCode{
+		return &model.ErrorWithStatusCode{
 			Error:      textResponse.Error,
 			StatusCode: resp.StatusCode,
 		}, nil
@@ -133,9 +138,9 @@ func Handler(c *gin.Context, resp *http.Response, promptTokens int, model string
 	if textResponse.Usage.TotalTokens == 0 {
 		completionTokens := 0
 		for _, choice := range textResponse.Choices {
-			completionTokens += CountTokenText(choice.Message.StringContent(), model)
+			completionTokens += CountTokenText(choice.Message.StringContent(), modelName)
 		}
-		textResponse.Usage = Usage{
+		textResponse.Usage = model.Usage{
 			PromptTokens:     promptTokens,
 			CompletionTokens: completionTokens,
 			TotalTokens:      promptTokens + completionTokens,

relay/channel/openai/model.go

@@ -1,15 +1,6 @@
 package openai
 
-type Message struct {
-	Role    string  `json:"role"`
-	Content any     `json:"content"`
-	Name    *string `json:"name,omitempty"`
-}
-
-type ImageURL struct {
-	Url    string `json:"url,omitempty"`
-	Detail string `json:"detail,omitempty"`
-}
+import "github.com/songquanpeng/one-api/relay/model"
 
 type TextContent struct {
 	Type string `json:"type,omitempty"`
@@ -17,142 +8,21 @@ type TextContent struct {
 }
 
 type ImageContent struct {
-	Type     string    `json:"type,omitempty"`
-	ImageURL *ImageURL `json:"image_url,omitempty"`
-}
-
-type OpenAIMessageContent struct {
-	Type     string    `json:"type,omitempty"`
-	Text     string    `json:"text"`
-	ImageURL *ImageURL `json:"image_url,omitempty"`
-}
-
-func (m Message) IsStringContent() bool {
-	_, ok := m.Content.(string)
-	return ok
-}
-
-func (m Message) StringContent() string {
-	content, ok := m.Content.(string)
-	if ok {
-		return content
-	}
-	contentList, ok := m.Content.([]any)
-	if ok {
-		var contentStr string
-		for _, contentItem := range contentList {
-			contentMap, ok := contentItem.(map[string]any)
-			if !ok {
-				continue
-			}
-			if contentMap["type"] == ContentTypeText {
-				if subStr, ok := contentMap["text"].(string); ok {
-					contentStr += subStr
-				}
-			}
-		}
-		return contentStr
-	}
-	return ""
-}
-
-func (m Message) ParseContent() []OpenAIMessageContent {
-	var contentList []OpenAIMessageContent
-	content, ok := m.Content.(string)
-	if ok {
-		contentList = append(contentList, OpenAIMessageContent{
-			Type: ContentTypeText,
-			Text: content,
-		})
-		return contentList
-	}
-	anyList, ok := m.Content.([]any)
-	if ok {
-		for _, contentItem := range anyList {
-			contentMap, ok := contentItem.(map[string]any)
-			if !ok {
-				continue
-			}
-			switch contentMap["type"] {
-			case ContentTypeText:
-				if subStr, ok := contentMap["text"].(string); ok {
-					contentList = append(contentList, OpenAIMessageContent{
-						Type: ContentTypeText,
-						Text: subStr,
-					})
-				}
-			case ContentTypeImageURL:
-				if subObj, ok := contentMap["image_url"].(map[string]any); ok {
-					contentList = append(contentList, OpenAIMessageContent{
-						Type: ContentTypeImageURL,
-						ImageURL: &ImageURL{
-							Url: subObj["url"].(string),
-						},
-					})
-				}
-			}
-		}
-		return contentList
-	}
-	return nil
-}
-
-type ResponseFormat struct {
-	Type string `json:"type,omitempty"`
-}
-
-type GeneralOpenAIRequest struct {
-	Model            string          `json:"model,omitempty"`
-	Messages         []Message       `json:"messages,omitempty"`
-	Prompt           any             `json:"prompt,omitempty"`
-	Stream           bool            `json:"stream,omitempty"`
-	MaxTokens        int             `json:"max_tokens,omitempty"`
-	Temperature      float64         `json:"temperature,omitempty"`
-	TopP             float64         `json:"top_p,omitempty"`
-	N                int             `json:"n,omitempty"`
-	Input            any             `json:"input,omitempty"`
-	Instruction      string          `json:"instruction,omitempty"`
-	Size             string          `json:"size,omitempty"`
-	Functions        any             `json:"functions,omitempty"`
-	FrequencyPenalty float64         `json:"frequency_penalty,omitempty"`
-	PresencePenalty  float64         `json:"presence_penalty,omitempty"`
-	ResponseFormat   *ResponseFormat `json:"response_format,omitempty"`
-	Seed             float64         `json:"seed,omitempty"`
-	Tools            any             `json:"tools,omitempty"`
-	ToolChoice       any             `json:"tool_choice,omitempty"`
-	User             string          `json:"user,omitempty"`
-}
-
-func (r GeneralOpenAIRequest) ParseInput() []string {
-	if r.Input == nil {
-		return nil
-	}
-	var input []string
-	switch r.Input.(type) {
-	case string:
-		input = []string{r.Input.(string)}
-	case []any:
-		input = make([]string, 0, len(r.Input.([]any)))
-		for _, item := range r.Input.([]any) {
-			if str, ok := item.(string); ok {
-				input = append(input, str)
-			}
-		}
-	}
-	return input
+	Type     string          `json:"type,omitempty"`
+	ImageURL *model.ImageURL `json:"image_url,omitempty"`
 }
 
 type ChatRequest struct {
-	Model     string    `json:"model"`
-	Messages  []Message `json:"messages"`
-	MaxTokens int       `json:"max_tokens"`
+	Model     string          `json:"model"`
+	Messages  []model.Message `json:"messages"`
+	MaxTokens int             `json:"max_tokens"`
 }
 
 type TextRequest struct {
-	Model     string    `json:"model"`
-	Messages  []Message `json:"messages"`
-	Prompt    string    `json:"prompt"`
-	MaxTokens int       `json:"max_tokens"`
+	Model     string          `json:"model"`
+	Messages  []model.Message `json:"messages"`
+	Prompt    string          `json:"prompt"`
+	MaxTokens int             `json:"max_tokens"`
 	//Stream   bool      `json:"stream"`
 }
 
@@ -201,48 +71,30 @@ type TextToSpeechRequest struct {
 	ResponseFormat string  `json:"response_format"`
 }
 
-type Usage struct {
-	PromptTokens     int `json:"prompt_tokens"`
-	CompletionTokens int `json:"completion_tokens"`
-	TotalTokens      int `json:"total_tokens"`
-}
-
 type UsageOrResponseText struct {
-	*Usage
+	*model.Usage
 	ResponseText string
 }
 
-type Error struct {
-	Message string `json:"message"`
-	Type    string `json:"type"`
-	Param   string `json:"param"`
-	Code    any    `json:"code"`
-}
-
-type ErrorWithStatusCode struct {
-	Error
-	StatusCode int `json:"status_code"`
-}
-
 type SlimTextResponse struct {
-	Choices []TextResponseChoice `json:"choices"`
-	Usage   `json:"usage"`
-	Error   Error `json:"error"`
+	Choices     []TextResponseChoice `json:"choices"`
+	model.Usage `json:"usage"`
+	Error       model.Error `json:"error"`
 }
 
 type TextResponseChoice struct {
-	Index        int `json:"index"`
-	Message      `json:"message"`
-	FinishReason string `json:"finish_reason"`
+	Index         int `json:"index"`
+	model.Message `json:"message"`
+	FinishReason  string `json:"finish_reason"`
 }
 
 type TextResponse struct {
-	Id      string               `json:"id"`
-	Model   string               `json:"model,omitempty"`
-	Object  string               `json:"object"`
-	Created int64                `json:"created"`
-	Choices []TextResponseChoice `json:"choices"`
-	Usage   `json:"usage"`
+	Id          string               `json:"id"`
+	Model       string               `json:"model,omitempty"`
+	Object      string               `json:"object"`
+	Created     int64                `json:"created"`
+	Choices     []TextResponseChoice `json:"choices"`
+	model.Usage `json:"usage"`
 }
 
 type EmbeddingResponseItem struct {
@@ -252,10 +104,10 @@ type EmbeddingResponseItem struct {
 }
 
 type EmbeddingResponse struct {
-	Object string                  `json:"object"`
-	Data   []EmbeddingResponseItem `json:"data"`
-	Model  string                  `json:"model"`
-	Usage  `json:"usage"`
+	Object      string                  `json:"object"`
+	Data        []EmbeddingResponseItem `json:"data"`
+	Model       string                  `json:"model"`
+	model.Usage `json:"usage"`
 }
 
 type ImageResponse struct {
@@ -266,8 +118,10 @@ type ImageResponse struct {
 }
 
 type ChatCompletionsStreamResponseChoice struct {
+	Index int `json:"index"`
 	Delta struct {
 		Content string `json:"content"`
+		Role    string `json:"role,omitempty"`
 	} `json:"delta"`
 	FinishReason *string `json:"finish_reason,omitempty"`
 }
@@ -278,6 +132,7 @@ type ChatCompletionsStreamResponse struct {
 	Created int64                                 `json:"created"`
 	Model   string                                `json:"model"`
 	Choices []ChatCompletionsStreamResponseChoice `json:"choices"`
+	Usage   *model.Usage                          `json:"usage"`
 }
 
 type CompletionsStreamResponse struct {

relay/channel/openai/token.go

@@ -4,11 +4,12 @@ import (
 	"errors"
 	"fmt"
 	"github.com/pkoukk/tiktoken-go"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/config"
+	"github.com/songquanpeng/one-api/common/image"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/model"
 	"math"
-	"one-api/common"
-	"one-api/common/config"
-	"one-api/common/image"
-	"one-api/common/logger"
 	"strings"
 )
 
@@ -27,7 +28,7 @@ func InitTokenEncoders() {
 	if err != nil {
 		logger.FatalLog(fmt.Sprintf("failed to get gpt-4 token encoder: %s", err.Error()))
 	}
-	for model, _ := range common.ModelRatio {
+	for model := range common.ModelRatio {
 		if strings.HasPrefix(model, "gpt-3.5") {
 			tokenEncoderMap[model] = gpt35TokenEncoder
 		} else if strings.HasPrefix(model, "gpt-4") {
@@ -63,7 +64,7 @@ func getTokenNum(tokenEncoder *tiktoken.Tiktoken, text string) int {
 	return len(tokenEncoder.Encode(text, nil, nil))
 }
 
-func CountTokenMessages(messages []Message, model string) int {
+func CountTokenMessages(messages []model.Message, model string) int {
 	tokenEncoder := getTokenEncoder(model)
 	// Reference:
 	// https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb

relay/channel/openai/util.go

@@ -1,12 +1,14 @@
 package openai
 
-func ErrorWrapper(err error, code string, statusCode int) *ErrorWithStatusCode {
-	Error := Error{
+import "github.com/songquanpeng/one-api/relay/model"
+
+func ErrorWrapper(err error, code string, statusCode int) *model.ErrorWithStatusCode {
+	Error := model.Error{
 		Message: err.Error(),
 		Type:    "one_api_error",
 		Code:    code,
 	}
-	return &ErrorWithStatusCode{
+	return &model.ErrorWithStatusCode{
 		Error:      Error,
 		StatusCode: statusCode,
 	}

relay/channel/palm/adaptor.go

@@ -0,0 +1,60 @@
+package palm
+
+import (
+	"errors"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
+	"net/http"
+)
+
+type Adaptor struct {
+}
+
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	return fmt.Sprintf("%s/v1beta2/models/chat-bison-001:generateMessage", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("x-goog-api-key", meta.APIKey)
+	return nil
+}
+
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return ConvertRequest(*request), nil
+}
+
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		var responseText string
+		err, responseText = StreamHandler(c, resp)
+		usage = openai.ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+	} else {
+		err, usage = Handler(c, resp, meta.PromptTokens, meta.ActualModelName)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "google palm"
+}

relay/channel/palm/constants.go

@@ -0,0 +1,5 @@
+package palm
+
+var ModelList = []string{
+	"PaLM-2",
+}

relay/channel/palm/model.go

@@ -0,0 +1,40 @@
+package palm
+
+import (
+	"github.com/songquanpeng/one-api/relay/model"
+)
+
+type ChatMessage struct {
+	Author  string `json:"author"`
+	Content string `json:"content"`
+}
+
+type Filter struct {
+	Reason  string `json:"reason"`
+	Message string `json:"message"`
+}
+
+type Prompt struct {
+	Messages []ChatMessage `json:"messages"`
+}
+
+type ChatRequest struct {
+	Prompt         Prompt  `json:"prompt"`
+	Temperature    float64 `json:"temperature,omitempty"`
+	CandidateCount int     `json:"candidateCount,omitempty"`
+	TopP           float64 `json:"topP,omitempty"`
+	TopK           int     `json:"topK,omitempty"`
+}
+
+type Error struct {
+	Code    int    `json:"code"`
+	Message string `json:"message"`
+	Status  string `json:"status"`
+}
+
+type ChatResponse struct {
+	Candidates []ChatMessage   `json:"candidates"`
+	Messages   []model.Message `json:"messages"`
+	Filters    []Filter        `json:"filters"`
+	Error      Error           `json:"error"`
+}

relay/channel/palm/palm.go

@@ -1,25 +1,26 @@
-package google
+package palm
 
 import (
 	"encoding/json"
 	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/common"
+	"github.com/songquanpeng/one-api/common/helper"
+	"github.com/songquanpeng/one-api/common/logger"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/constant"
+	"github.com/songquanpeng/one-api/relay/model"
 	"io"
 	"net/http"
-	"one-api/common"
-	"one-api/common/helper"
-	"one-api/common/logger"
-	"one-api/relay/channel/openai"
-	"one-api/relay/constant"
 )
 
 // https://developers.generativeai.google/api/rest/generativelanguage/models/generateMessage#request-body
 // https://developers.generativeai.google/api/rest/generativelanguage/models/generateMessage#response-body
 
-func ConvertPaLMRequest(textRequest openai.GeneralOpenAIRequest) *PaLMChatRequest {
-	palmRequest := PaLMChatRequest{
-		Prompt: PaLMPrompt{
-			Messages: make([]PaLMChatMessage, 0, len(textRequest.Messages)),
+func ConvertRequest(textRequest model.GeneralOpenAIRequest) *ChatRequest {
+	palmRequest := ChatRequest{
+		Prompt: Prompt{
+			Messages: make([]ChatMessage, 0, len(textRequest.Messages)),
 		},
 		Temperature:    textRequest.Temperature,
 		CandidateCount: textRequest.N,
@@ -27,7 +28,7 @@ func ConvertPaLMRequest(textRequest openai.GeneralOpenAIRequest) *PaLMChatReques
 		TopK:           textRequest.MaxTokens,
 	}
 	for _, message := range textRequest.Messages {
-		palmMessage := PaLMChatMessage{
+		palmMessage := ChatMessage{
 			Content: message.StringContent(),
 		}
 		if message.Role == "user" {
@@ -40,14 +41,14 @@ func ConvertPaLMRequest(textRequest openai.GeneralOpenAIRequest) *PaLMChatReques
 	return &palmRequest
 }
 
-func responsePaLM2OpenAI(response *PaLMChatResponse) *openai.TextResponse {
+func responsePaLM2OpenAI(response *ChatResponse) *openai.TextResponse {
 	fullTextResponse := openai.TextResponse{
 		Choices: make([]openai.TextResponseChoice, 0, len(response.Candidates)),
 	}
 	for i, candidate := range response.Candidates {
 		choice := openai.TextResponseChoice{
 			Index: i,
-			Message: openai.Message{
+			Message: model.Message{
 				Role:    "assistant",
 				Content: candidate.Content,
 			},
@@ -58,7 +59,7 @@ func responsePaLM2OpenAI(response *PaLMChatResponse) *openai.TextResponse {
 	return &fullTextResponse
 }
 
-func streamResponsePaLM2OpenAI(palmResponse *PaLMChatResponse) *openai.ChatCompletionsStreamResponse {
+func streamResponsePaLM2OpenAI(palmResponse *ChatResponse) *openai.ChatCompletionsStreamResponse {
 	var choice openai.ChatCompletionsStreamResponseChoice
 	if len(palmResponse.Candidates) > 0 {
 		choice.Delta.Content = palmResponse.Candidates[0].Content
@@ -71,7 +72,7 @@ func streamResponsePaLM2OpenAI(palmResponse *PaLMChatResponse) *openai.ChatCompl
 	return &response
 }
 
-func PaLMStreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, string) {
+func StreamHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, string) {
 	responseText := ""
 	responseId := fmt.Sprintf("chatcmpl-%s", helper.GetUUID())
 	createdTime := helper.GetTimestamp()
@@ -90,7 +91,7 @@ func PaLMStreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithSt
 			stopChan <- true
 			return
 		}
-		var palmResponse PaLMChatResponse
+		var palmResponse ChatResponse
 		err = json.Unmarshal(responseBody, &palmResponse)
 		if err != nil {
 			logger.SysError("error unmarshalling stream response: " + err.Error())
@@ -130,7 +131,7 @@ func PaLMStreamHandler(c *gin.Context, resp *http.Response) (*openai.ErrorWithSt
 	return nil, responseText
 }
 
-func PaLMHandler(c *gin.Context, resp *http.Response, promptTokens int, model string) (*openai.ErrorWithStatusCode, *openai.Usage) {
+func Handler(c *gin.Context, resp *http.Response, promptTokens int, modelName string) (*model.ErrorWithStatusCode, *model.Usage) {
 	responseBody, err := io.ReadAll(resp.Body)
 	if err != nil {
 		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
@@ -139,14 +140,14 @@ func PaLMHandler(c *gin.Context, resp *http.Response, promptTokens int, model st
 	if err != nil {
 		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
 	}
-	var palmResponse PaLMChatResponse
+	var palmResponse ChatResponse
 	err = json.Unmarshal(responseBody, &palmResponse)
 	if err != nil {
 		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
 	}
 	if palmResponse.Error.Code != 0 || len(palmResponse.Candidates) == 0 {
-		return &openai.ErrorWithStatusCode{
-			Error: openai.Error{
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
 				Message: palmResponse.Error.Message,
 				Type:    palmResponse.Error.Status,
 				Param:   "",
@@ -156,9 +157,9 @@ func PaLMHandler(c *gin.Context, resp *http.Response, promptTokens int, model st
 		}, nil
 	}
 	fullTextResponse := responsePaLM2OpenAI(&palmResponse)
-	fullTextResponse.Model = model
-	completionTokens := openai.CountTokenText(palmResponse.Candidates[0].Content, model)
-	usage := openai.Usage{
+	fullTextResponse.Model = modelName
+	completionTokens := openai.CountTokenText(palmResponse.Candidates[0].Content, modelName)
+	usage := model.Usage{
 		PromptTokens:     promptTokens,
 		CompletionTokens: completionTokens,
 		TotalTokens:      promptTokens + completionTokens,

relay/channel/tencent/adaptor.go

@@ -1,22 +1,76 @@
 package tencent
 
 import (
+	"errors"
+	"fmt"
 	"github.com/gin-gonic/gin"
+	"github.com/songquanpeng/one-api/relay/channel"
+	"github.com/songquanpeng/one-api/relay/channel/openai"
+	"github.com/songquanpeng/one-api/relay/model"
+	"github.com/songquanpeng/one-api/relay/util"
+	"io"
 	"net/http"
-	"one-api/relay/channel/openai"
+	"strings"
 )
 
+// https://cloud.tencent.com/document/api/1729/101837
+
 type Adaptor struct {
+	Sign string
 }
 
-func (a *Adaptor) Auth(c *gin.Context) error {
+func (a *Adaptor) Init(meta *util.RelayMeta) {
+
+}
+
+func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
+	return fmt.Sprintf("%s/hyllm/v1/chat/completions", meta.BaseURL), nil
+}
+
+func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
+	channel.SetupCommonRequestHeader(c, req, meta)
+	req.Header.Set("Authorization", a.Sign)
+	req.Header.Set("X-TC-Action", meta.ActualModelName)
 	return nil
 }
 
-func (a *Adaptor) ConvertRequest(request *openai.GeneralOpenAIRequest) (any, error) {
-	return nil, nil
+func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.GeneralOpenAIRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	apiKey := c.Request.Header.Get("Authorization")
+	apiKey = strings.TrimPrefix(apiKey, "Bearer ")
+	appId, secretId, secretKey, err := ParseConfig(apiKey)
+	if err != nil {
+		return nil, err
+	}
+	tencentRequest := ConvertRequest(*request)
+	tencentRequest.AppId = appId
+	tencentRequest.SecretId = secretId
+	// we have to calculate the sign here
+	a.Sign = GetSign(*tencentRequest, secretKey)
+	return tencentRequest, nil
 }
 
-func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response) (*openai.ErrorWithStatusCode, *openai.Usage, error) {
-	return nil, nil, nil
+func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
+	return channel.DoRequestHelper(a, c, meta, requestBody)
+}
+
+func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.RelayMeta) (usage *model.Usage, err *model.ErrorWithStatusCode) {
+	if meta.IsStream {
+		var responseText string
+		err, responseText = StreamHandler(c, resp)
+		usage = openai.ResponseText2Usage(responseText, meta.ActualModelName, meta.PromptTokens)
+	} else {
+		err, usage = Handler(c, resp)
+	}
+	return
+}
+
+func (a *Adaptor) GetModelList() []string {
+	return ModelList
+}
+
+func (a *Adaptor) GetChannelName() string {
+	return "tencent"
 }

relay/channel/tencent/constants.go

@@ -0,0 +1,7 @@
+package tencent
+
+var ModelList = []string{
+	"ChatPro",
+	"ChatStd",
+	"hunyuan",
+}